PDA

View Full Version : Chinese GEDmatch averages



okarinaofsteiner
07-29-2021, 01:10 AM
Creating a separate thread for this from my East Eurasian megathread (https://anthrogenica.com/showthread.php?20657-okarinaofsteiner-s-East-Eurasian-GEDmatch-megathread&p=783117&viewfull=1#post783117): https://anthrogenica.com/showthread.php?20657-okarinaofsteiner-s-East-Eurasian-GEDmatch-megathread

Anthroscape user @kushkush made some autosomal DNA (HarappaWorld) maps of Chinese provinces around the end of 2019, which were interesting but didn't seem completely accurate to me, since they weren't weighted by population distribution within specific provinces. I think these are just averages of random samples from specific provinces, they don't seem to be specific to Han Chinese.

Posting their findings here for reference:

HarappaWorld NE Asian
https://i.imgur.com/XXiQ5jM.png

HarappaWorld SE Asian
https://i.imgur.com/AlrCqsi.png

HarappaWorld Siberian
https://i.imgur.com/e0vq3KL.png

HarappaWorld table
https://i.imgur.com/Rg3PgpF.png

okarinaofsteiner
07-29-2021, 01:14 AM
Reposting my own version of @kushkush's maps with MDLP K23b results (https://anthrogenica.com/showthread.php?20657-okarinaofsteiner-s-East-Eurasian-GEDmatch-megathread&p=783117&viewfull=1#post783117) here for comparison.

This was based on the DNAConnect.org dataset as well as other samples found on GEDmatch and the WeGene forum. I'd like to think this is more representative of the Han in each province, since I tried to weigh my assigned samples for each province I had more than 10 samples for according to the population distibution for that province.

The first 4 maps are of the 4 largest MDLP K23b East Eurasian ancestry components found among Chinese samples. The 5th one is of a North-South cline I created for East Asian MDLP K23b results. Basically "Tungus_Altaic" = northern, "Austronesian" = southern, while "South_East_Asian" and the 2 "Siberian" components are defined relative to T_A and AN. The original idea was that a Han Chinese person who scores the same amount of T_A and AN will score 0.5 on my North-South cline.

https://i.imgur.com/cw2QvTF.png

https://i.imgur.com/hwr2amo.png

https://i.imgur.com/rT227Kd.png

https://i.imgur.com/IbsfvhK.png

https://i.imgur.com/EQjAhu7.png

Original methodology post (link is defunct)- https://www.tapatalk.com/groups/anthroscape/mdlp-k23b-results-and-pca-plots-for-east-asians-t85446-s120.html#p1848187

TL;DR- I tried to weigh my available data according to the actual population distribution within each province (the DNAConnect.org data was geographically biased for most of the provinces- especially Jiangxi, Guangdong, Hunan, and Chongqing- which make up most of the DNAConnect.org dataset). Most of the samples in my original dataset and from DNAConnect.org were from the southern provinces, with very few from the northern provinces, so I had to use whatever I could get my hands on. The samples I used in my original dataset were just the ones of known regional/provincial ancestry.

I weighed the Fujian, Jiangxi, Hunan, and Anhui DNAConnect.org subsets according to the population distribution of each province to make them more representative. The Guangdong and Shaanxi subsets only represented more remote parts of the provinces, and the Guizhou subset was mostly from minority-heavy areas. The only northern province represented in the DNAConnect.org dataset was Henan.

Kaazi
07-29-2021, 04:46 AM
Creating a separate thread for this from my East Eurasian megathread (https://anthrogenica.com/showthread.php?20657-okarinaofsteiner-s-East-Eurasian-GEDmatch-megathread&p=783117&viewfull=1#post783117): https://anthrogenica.com/showthread.php?20657-okarinaofsteiner-s-East-Eurasian-GEDmatch-megathread

Anthroscape user @kushkush made some autosomal DNA (HarappaWorld) maps of Chinese provinces around the end of 2019, which were interesting but didn't seem completely accurate to me, since they weren't weighted by population distribution within specific provinces. I think these are just averages of random samples from specific provinces, they don't seem to be specific to Han Chinese.

Posting their findings here for reference:

HarappaWorld NE Asian
https://i.imgur.com/XXiQ5jM.png

HarappaWorld SE Asian
https://i.imgur.com/AlrCqsi.png

HarappaWorld Siberian
https://i.imgur.com/e0vq3KL.png

HarappaWorld table
https://i.imgur.com/Rg3PgpF.png

That's pretty outdated table.

You should create new average based on the large no. of Tibetan, Han samples available in genoplot. Tibet surely isn't 78% NE-Asian on average. It would be sth around early 70s.


SampleNE-AsianPapuanSE-AsianS-IndianSiberianSW-AsianBeringianSanNE-EuroAmericanCaucasianBalochMediterraneanE-AfricanPygmyW-AfricanTibetan Xinlong ► XL7699.760.180.060000000000000
Tibetan Xinlong ► XL6080.430.6207.4111.20000000000.030.31
Tibetan Yajiang ► YJ14080.031.051.324.311.060.010.4001.130000.490.21 0
Tibetan Chamdo ► T27479.971.7205.1410.640.541.020.1000.8600000
Tibetan Yajiang ► YJ11179.81.382.573.7111.140000.2300.0400.71000.4
Tibetan Yajiang ► YJ3779.32.5603.9411.810.490001.50.250000.140
Tibetan Chamdo ► T24778.991.090.347.8711.290.410000000000
Tibetan Chamdo ► T7478.421.1406.5813.300.55000000000
Tibetan Yajiang ► YJ0178.310.421.594.0212.5900.45000.86000.510.3200. 94
Tibetan Yajiang ► YJ2578.140.310.935.0411.540.551.990.14000.1700.020 .1101.06
Tibetan Yajiang ► YJ10478.012.13.023.8812.010.03000.4300000.5100
Tibetan Chamdo ► T28377.911.071.225.4312.2800.810.28000.9300.060.01 00
Tibetan Chamdo ► T27377.661.2914.8711.040100.591.230.950.370000
Tibetan Lhasa ► T7677.641.1905.7111.2101.990000.841.060.35000
Tibetan Chamdo ► T19477.642.212.494.512.5800.05000.17000.36000
Tibetan Chamdo ► T17377.581.2806.6412.4901.840.1700000000
Tibetan Xinlong ► XL0577.420.8808.312.810.18000.120.29000000
Tibetan Chamdo ► T14477.421.9105.3812.5500.840010.900000
Tibetan Chamdo ► T19677.412.172.72.9511.230.22.9800.310.05000000
Tibetan Nagqu ► T23877.341.060.94.1612.490.591.840.081.530000000
Tibetan Nagqu ► T32077.261.971.14.6511.960.490.5701.4600.010.3100. 2200
Tibetan Chamdo ► T14177.151.090.715.9713.5600.3600.750.43000000
Tibetan Yajiang ► YJ1277.010.626.456.346.98000.320.261.2500.030.310. 4300
Tibetan Lhasa ► T18576.4710.198.3411.7600000.720.491.030000
Tibetan Chamdo ► T36376.261.660.595.813.5900.51000.1301.14000.340
Tibetan Lhasa ► T16576.210.8808.1111.301.410002.0900000
Tibetan Shigatse ► T8076.182.120.798.289.640.510.61001.1400.5400.1800
Tibetan Nagqu ► T6776.092.11.436.7511.7100.530.170.4700.7600000
Tibetan Nagqu ► T12175.651.6507.8711.80.481.03001.0500.100.3600
Tibetan Lhasa ► T18375.431.191.425.613.04000002.130.960.22000
Tibetan Nagqu ► T9175.271.3806.8113.550.310.770.3600.230.93000.410 0
Tibetan Shannan ► T12775.242.70.257.8311.0600.530.0600.940.360.660.3 5000
Tibetan Shigatse ► T2275.082.031.297.0211.710000.021.181.310.370000
Tibetan Nagqu ► T31274.921.1909.3114.060000.30.0600.170000
Tibetan Chamdo ► T28274.921.980.526.212.800.08001.241.7800.360.0900
Tibetan Shigatse ► T21574.71.830.927.4412.300.9800.1700.61.050000
Tibetan Shannan ► T28974.592.1608.2610.9800.9400.890.3401.840000
Tibetan Xinlong ► XL6174.41.0710.964.288.25000.1200.840000.050.020
Tibetan Lhasa ► T17074.342.381.17.7910.811.521.060.090.440.28000.2 000
Tibetan Shannan ► T20074.281.7907.9612.840.030000.7901.6800.6400
Tibetan Shannan ► T18874.052.010.178.5110.4600.7200.751.431.8200.060 00
Tibetan Shigatse ► T7873.822.860.596.0413.790.30.040.1400.650.541.210 000
Tibetan Lhasa ► T18173.781.361.566.9312.2400.7101.81.1800.440000
Tibetan Lhasa ► T12573.781.880.527.1411.240.492.92000.3801.660000
Tibetan Shannan ► T24273.62.920.766.2120.211.65000.970.70.75000.230
Tibetan Xinlong ► XL12073.531.590.436.1613.350.232.0100.130.930.860. 130.100.30.25
Tibetan Nagqu ► T18773.521.652.266.5613.1400.71000.6101.560000
Tibetan Shigatse ► T29773.52.560.846.513.5400.34000.6202.10000
Tibetan Gangcha ► QH473.351.7805.3912.6102.1601.020.6501.951.09000
Tibetan Lhasa ► T2173.231.412.17.4412.7301.390.1201.2600.320000
Tibetan Gangcha ► QH2073.210.860.34.6715.2100001.41.532.810000
Tibetan Shigatse ► T19272.681.172.997.0611.4901.8901.650.2800.4500.35 00
Tibetan Lhasa ► T1872.451.34.326.1112.741.560000.170.510.830000
Tibetan Yajiang ► YJ1472.291.0613.073.458.710.52000.050.230.1100000. 5
Tibetan Shigatse ► T291721.681.957.6313.8100000.7801.3700.7800
Tibetan Shigatse ► T7771.921.791.158.3311.3101.2901.511.3101.410000
Tibetan Xunhua ► XHTB771.821.155.63.5410.732.2100.041.800.072.45000 .210.38
Tibetan Gangcha ► QH1171.811.862.314.1914.60.560.450.111.1901.030.90 00
Tibetan Gangcha ► QH171.711.730.545.9712.270.91001.432.30.262.890000
Tibetan Gangcha ► QH571.70.380.795.2514.513.610.180.181.121.30.180.3 90.3300.070
Tibetan Shigatse ► T4171.482.321.168.1113.53000.030.612.11000.66000
Tibetan Shannan ► T19071.281.893.036.8813.030.7800.110.731.261.01000 00
Tibetan Gangcha ► QH1771.241.760.265.1214.370.922.5700.6301.980.320. 83000
Tibetan Gangcha ► QH1271.160.483.154.7115.4401.0502.021.370.6200000
Tibetan Nagqu ► T27871.041.743.517.6513.330.640.4700.061.550000.01 00
Tibetan Shannan ► T24870.551.773.617.612.470.91.260.0800.8200.930000
Tibetan Gangcha ► QH870.52.560.185.3814.301.7202.6200.31.930.51000
Tibetan Shigatse ► T18970.22.243.589.1913.410.010.13000.30.510.31000. 110
Tibetan Shannan ► T19970.141.932.187.7212.3302.0500.060.630.432.1000 .430
Tibetan Gangcha ► QH1070.11.560.884.6314.6700011.121.513.320.96000.2 6
Tibetan Yajiang ► YJ3969.980.9814.613.368.970000.881.09000000.13
Tibetan Gangcha ► QH1969.880.874.344.9615.1800.4700.6102.830.860000
Tibetan Gangcha ► QH1869.871.50.514.7415.40.481.0801.080.913.291.130 000
Tibetan Gangcha ► QH669.732.173.224.0514.050.932.8200.1600.421.211.2 3000
Tibetan Xinlong ► XL4769.48014.484.138.4700.980.150.9400.330.750.050 00.25
Tibetan Gangcha ► QH1569.481.013.026.515.420000.350.932.500.300.490
Tibetan Shannan ► T12669.351.542.816.1115.250.630.9800.2202.470.5100 00.13
Tibetan Xinlong ► XL9268.9021.660.817.330.280.570000.1800000.25
Tibetan Xunhua ► XHTB2068.851.644.196.3510.6800.980.523.17000.820.2 9002.53
Tibetan Gangcha ► QH1468.751.982.512.6816.4701.8300.9503.141.150.530 00
Tibetan Gangcha ► QH768.450.72.83.3816.4601.810.161.0800.962.51.6800 0
Tibetan Xinlong ► XL3768.270.7424.880.195.1400.78000000000
Tibetan Gannan ► GN1968.1907.23.3710.420.61.970.061.080.783.331.910 .4300.650
Tibetan Xunhua ► XHTB568.071.165.416.8412.3901.160.230.931.121.830. 79000.060
Tibetan Gannan ► GN0168.020.449.072.189.95000.4301.371.494.722.3400 0
Tibetan Gangcha ► QH1667.941.362.257.1213.2201.0100.091.582.151.511. 76000
Tibetan Gangcha ► QH967.760.965.323.8615.5500.350.141.7400.913.41000 0
Tibetan Gangcha ► QH1367.30.3411.232.2410.010.721.5401.411.982.330.7 800.1300
Tibetan Gannan ► GN1067.18012.150.8110.1300.803.0101.792.821.10.200
Tibetan Xinlong ► XL2266.940.727.690.052.5101.520.1100.34000000.14
Tibetan Gannan ► GN1765.980.3712.023.028.780.51.2103.6402.571.050.8 7000
Tibetan Gannan ► GN0365.580.9512.131.768.851.190.301.2702.085.040.1 30.4100.31
Tibetan Gangcha ► QH364.671.422.65.9217.520.531.7302.70.940.870.660. 45000
Tibetan Xunhua ► XHTB1763.941.3814.61010.790.430.0401.1701.42.791.7 40.501.22
Tibetan Yajiang ► YJ8663.290.6612.512.257.161.191.450.081.170.616.87 2.760000
Tibetan Gangcha ► QH262.371.561.724.4318.981.161.760.020.961.911.683 .460000
Tibetan Xinlong ► XL2059.360.1736.450.172.490.8800.100.140.12000.120 0

Shuzam87
08-05-2021, 12:30 AM
That's pretty outdated table.

You should create new average based on the large no. of Tibetan, Han samples available in genoplot. Tibet surely isn't 78% NE-Asian on average. It would be sth around early 70s.


SampleNE-AsianPapuanSE-AsianS-IndianSiberianSW-AsianBeringianSanNE-EuroAmericanCaucasianBalochMediterraneanE-AfricanPygmyW-AfricanTibetan Xinlong ► XL7699.760.180.060000000000000
Tibetan Xinlong ► XL6080.430.6207.4111.20000000000.030.31
Tibetan Yajiang ► YJ14080.031.051.324.311.060.010.4001.130000.490.21 0
Tibetan Chamdo ► T27479.971.7205.1410.640.541.020.1000.8600000
Tibetan Yajiang ► YJ11179.81.382.573.7111.140000.2300.0400.71000.4
Tibetan Yajiang ► YJ3779.32.5603.9411.810.490001.50.250000.140
Tibetan Chamdo ► T24778.991.090.347.8711.290.410000000000
Tibetan Chamdo ► T7478.421.1406.5813.300.55000000000
Tibetan Yajiang ► YJ0178.310.421.594.0212.5900.45000.86000.510.3200. 94
Tibetan Yajiang ► YJ2578.140.310.935.0411.540.551.990.14000.1700.020 .1101.06
Tibetan Yajiang ► YJ10478.012.13.023.8812.010.03000.4300000.5100
Tibetan Chamdo ► T28377.911.071.225.4312.2800.810.28000.9300.060.01 00
Tibetan Chamdo ► T27377.661.2914.8711.040100.591.230.950.370000
Tibetan Lhasa ► T7677.641.1905.7111.2101.990000.841.060.35000
Tibetan Chamdo ► T19477.642.212.494.512.5800.05000.17000.36000
Tibetan Chamdo ► T17377.581.2806.6412.4901.840.1700000000
Tibetan Xinlong ► XL0577.420.8808.312.810.18000.120.29000000
Tibetan Chamdo ► T14477.421.9105.3812.5500.840010.900000
Tibetan Chamdo ► T19677.412.172.72.9511.230.22.9800.310.05000000
Tibetan Nagqu ► T23877.341.060.94.1612.490.591.840.081.530000000
Tibetan Nagqu ► T32077.261.971.14.6511.960.490.5701.4600.010.3100. 2200
Tibetan Chamdo ► T14177.151.090.715.9713.5600.3600.750.43000000
Tibetan Yajiang ► YJ1277.010.626.456.346.98000.320.261.2500.030.310. 4300
Tibetan Lhasa ► T18576.4710.198.3411.7600000.720.491.030000
Tibetan Chamdo ► T36376.261.660.595.813.5900.51000.1301.14000.340
Tibetan Lhasa ► T16576.210.8808.1111.301.410002.0900000
Tibetan Shigatse ► T8076.182.120.798.289.640.510.61001.1400.5400.1800
Tibetan Nagqu ► T6776.092.11.436.7511.7100.530.170.4700.7600000
Tibetan Nagqu ► T12175.651.6507.8711.80.481.03001.0500.100.3600
Tibetan Lhasa ► T18375.431.191.425.613.04000002.130.960.22000
Tibetan Nagqu ► T9175.271.3806.8113.550.310.770.3600.230.93000.410 0
Tibetan Shannan ► T12775.242.70.257.8311.0600.530.0600.940.360.660.3 5000
Tibetan Shigatse ► T2275.082.031.297.0211.710000.021.181.310.370000
Tibetan Nagqu ► T31274.921.1909.3114.060000.30.0600.170000
Tibetan Chamdo ► T28274.921.980.526.212.800.08001.241.7800.360.0900
Tibetan Shigatse ► T21574.71.830.927.4412.300.9800.1700.61.050000
Tibetan Shannan ► T28974.592.1608.2610.9800.9400.890.3401.840000
Tibetan Xinlong ► XL6174.41.0710.964.288.25000.1200.840000.050.020
Tibetan Lhasa ► T17074.342.381.17.7910.811.521.060.090.440.28000.2 000
Tibetan Shannan ► T20074.281.7907.9612.840.030000.7901.6800.6400
Tibetan Shannan ► T18874.052.010.178.5110.4600.7200.751.431.8200.060 00
Tibetan Shigatse ► T7873.822.860.596.0413.790.30.040.1400.650.541.210 000
Tibetan Lhasa ► T18173.781.361.566.9312.2400.7101.81.1800.440000
Tibetan Lhasa ► T12573.781.880.527.1411.240.492.92000.3801.660000
Tibetan Shannan ► T24273.62.920.766.2120.211.65000.970.70.75000.230
Tibetan Xinlong ► XL12073.531.590.436.1613.350.232.0100.130.930.860. 130.100.30.25
Tibetan Nagqu ► T18773.521.652.266.5613.1400.71000.6101.560000
Tibetan Shigatse ► T29773.52.560.846.513.5400.34000.6202.10000
Tibetan Gangcha ► QH473.351.7805.3912.6102.1601.020.6501.951.09000
Tibetan Lhasa ► T2173.231.412.17.4412.7301.390.1201.2600.320000
Tibetan Gangcha ► QH2073.210.860.34.6715.2100001.41.532.810000
Tibetan Shigatse ► T19272.681.172.997.0611.4901.8901.650.2800.4500.35 00
Tibetan Lhasa ► T1872.451.34.326.1112.741.560000.170.510.830000
Tibetan Yajiang ► YJ1472.291.0613.073.458.710.52000.050.230.1100000. 5
Tibetan Shigatse ► T291721.681.957.6313.8100000.7801.3700.7800
Tibetan Shigatse ► T7771.921.791.158.3311.3101.2901.511.3101.410000
Tibetan Xunhua ► XHTB771.821.155.63.5410.732.2100.041.800.072.45000 .210.38
Tibetan Gangcha ► QH1171.811.862.314.1914.60.560.450.111.1901.030.90 00
Tibetan Gangcha ► QH171.711.730.545.9712.270.91001.432.30.262.890000
Tibetan Gangcha ► QH571.70.380.795.2514.513.610.180.181.121.30.180.3 90.3300.070
Tibetan Shigatse ► T4171.482.321.168.1113.53000.030.612.11000.66000
Tibetan Shannan ► T19071.281.893.036.8813.030.7800.110.731.261.01000 00
Tibetan Gangcha ► QH1771.241.760.265.1214.370.922.5700.6301.980.320. 83000
Tibetan Gangcha ► QH1271.160.483.154.7115.4401.0502.021.370.6200000
Tibetan Nagqu ► T27871.041.743.517.6513.330.640.4700.061.550000.01 00
Tibetan Shannan ► T24870.551.773.617.612.470.91.260.0800.8200.930000
Tibetan Gangcha ► QH870.52.560.185.3814.301.7202.6200.31.930.51000
Tibetan Shigatse ► T18970.22.243.589.1913.410.010.13000.30.510.31000. 110
Tibetan Shannan ► T19970.141.932.187.7212.3302.0500.060.630.432.1000 .430
Tibetan Gangcha ► QH1070.11.560.884.6314.6700011.121.513.320.96000.2 6
Tibetan Yajiang ► YJ3969.980.9814.613.368.970000.881.09000000.13
Tibetan Gangcha ► QH1969.880.874.344.9615.1800.4700.6102.830.860000
Tibetan Gangcha ► QH1869.871.50.514.7415.40.481.0801.080.913.291.130 000
Tibetan Gangcha ► QH669.732.173.224.0514.050.932.8200.1600.421.211.2 3000
Tibetan Xinlong ► XL4769.48014.484.138.4700.980.150.9400.330.750.050 00.25
Tibetan Gangcha ► QH1569.481.013.026.515.420000.350.932.500.300.490
Tibetan Shannan ► T12669.351.542.816.1115.250.630.9800.2202.470.5100 00.13
Tibetan Xinlong ► XL9268.9021.660.817.330.280.570000.1800000.25
Tibetan Xunhua ► XHTB2068.851.644.196.3510.6800.980.523.17000.820.2 9002.53
Tibetan Gangcha ► QH1468.751.982.512.6816.4701.8300.9503.141.150.530 00
Tibetan Gangcha ► QH768.450.72.83.3816.4601.810.161.0800.962.51.6800 0
Tibetan Xinlong ► XL3768.270.7424.880.195.1400.78000000000
Tibetan Gannan ► GN1968.1907.23.3710.420.61.970.061.080.783.331.910 .4300.650
Tibetan Xunhua ► XHTB568.071.165.416.8412.3901.160.230.931.121.830. 79000.060
Tibetan Gannan ► GN0168.020.449.072.189.95000.4301.371.494.722.3400 0
Tibetan Gangcha ► QH1667.941.362.257.1213.2201.0100.091.582.151.511. 76000
Tibetan Gangcha ► QH967.760.965.323.8615.5500.350.141.7400.913.41000 0
Tibetan Gangcha ► QH1367.30.3411.232.2410.010.721.5401.411.982.330.7 800.1300
Tibetan Gannan ► GN1067.18012.150.8110.1300.803.0101.792.821.10.200
Tibetan Xinlong ► XL2266.940.727.690.052.5101.520.1100.34000000.14
Tibetan Gannan ► GN1765.980.3712.023.028.780.51.2103.6402.571.050.8 7000
Tibetan Gannan ► GN0365.580.9512.131.768.851.190.301.2702.085.040.1 30.4100.31
Tibetan Gangcha ► QH364.671.422.65.9217.520.531.7302.70.940.870.660. 45000
Tibetan Xunhua ► XHTB1763.941.3814.61010.790.430.0401.1701.42.791.7 40.501.22
Tibetan Yajiang ► YJ8663.290.6612.512.257.161.191.450.081.170.616.87 2.760000
Tibetan Gangcha ► QH262.371.561.724.4318.981.161.760.020.961.911.683 .460000
Tibetan Xinlong ► XL2059.360.1736.450.172.490.8800.100.140.12000.120 0


Yeah, it would be cool if we can have an update.

okarinaofsteiner
08-10-2021, 11:54 PM
Yeah, it would be cool if we can have an update.

That was all Anthroscape user @kushkush- I have no idea if he's still into this stuff or have time for that. I don't even know how he found all those Tibetan GEDmatch samples back in the day.

I've considered making a HarappaWorld version of my MDLP K23b maps, but idk if the kits I used to calibrate results for different provinces are still publicly available. I'm guessing most have been deleted or removed from public access by now.

Kaazi
08-13-2021, 06:37 PM
Yeah, it would be cool if we can have an update.

Just made 100 Tibetans' average from genoplot.
https://anthrogenica.com/showthread.php?15471-South-Asian-HarappaWorld-results&p=791994&viewfull=1#post791994

https://i.imgur.com/SxAoANf.png

okarinaofsteiner
10-03-2021, 08:18 AM
A former Anthroscape member sent me a chart of 23mofang averages for various suburban and rural districts in China, which are believed to be more autosomally "representative" of specific regions and linguistic subgroups (as opposed to the city centers, which are more cosmopolitan and therefore more "mixed".) https://imgur.com/a/pxuhh3D

https://i.imgur.com/1UYN0Tn.jpg

Map (latitude + longitude) with ChinaMAP study clusters added
https://i.imgur.com/GX11i2q.png

I attempted to model the MDLP K23b results for these averages earlier this year by substituting the listed ancestry components with various combinations of MDLP K23b reference populations. For example, I would model "She" with MDLP K23b's She, "Tibetan" with the average of the 2 MDLP K23b "Tibetan" reference populations, etc.

Full model for all non-Han 23mofang components:

Daic (Tai-Kradai) = 5/16 "Chinese_Dai" + 1/16 part "Tai_Lue" + 2/16 parts "Jiamao" + 2/16 parts "Zhuang" + 5/16 parts "Dai" + 1/16 part "Ami_Taiwan"
Tungusic = average of: "Xibo" + "Oroqen" + "Hezhen" + "Daur"
Hmong-Mien = average of: "Yao" + "Miao" + "Hmong_Miao" + "Hmong"
Japanese = "Japanese" (the one that scores highest on T_A)
Korean = average of: "Korean_" + "Korean_KR"
Lahu = "Lahu"
Buryat = "Buryat"
Yakut = "Yakut" (doesn't include the other "Yakut" reference pop)
Uyghur = average of: "Uygur" + "Uygur-Han"
Tibeto-Burman = average of: "Tibetian_Madou" + "Tibetian_TTR"


This was difficult to do for the N_Han and S_Han components, because using HGDP's "Han_North" and "Han_" did not result in the Taiwanese, Hakka, and Cantonese area averages resembling the "Chinese_Taiwan", "Hakka", and "Cantonese" reference populations, so I did some tweaking.

N_Han = 46.17% S_EA, 33.55% T_A, 18.43% AN, based on "Han_North" (HGDP)
S_Han v1 = 4/10 Han_ (HGDP) + 2/10 Chinese_Dai + 1/10 Tujia + 1/10 She + 1/10 Korean (avg) + 1/10 Vietnamese
S_Han v2 = 4/14 Han_ (HGDP) + 2/14 Korean (avg) + 2/14 Vietnamese + 2/14 Chinese_Dai + 1/14 Tujia + 1/14 She + 1/14 Ami + 1/14 Naxi

The N_Han reference component was modeled as something slightly more "NE Asian" shifted than most actual Northern Han, since this was highest among the Shandong, Hebei, and southern Shanxi averages. The S_Han reference component was modeled as something in-between "Chinese_Taiwan"/"Hakka" and "Cantonese", since this component was highest among the Teochew, Leizhou, and Hakka-speaking area averages in Guangdong [the Yue-speaking areas scored lower on this component].

I decided to make another model for S_Han after seeing how my averages plotted on the Austronesian vs Tungus_Altaic graph against the Chinese samples from my original GEDmatch dataset. S_Han v2 matched the distribution better than S_Han v1.

Old version (S_Han v1)
https://i.imgur.com/w1LLi62.png

New version (S_Han v2)
https://i.imgur.com/SE1mMcf.png

MNOPSC1b
10-04-2021, 02:10 AM
Thanks for this awesome chart. We can see that Yue-speaking peoples indeed scored very low on Northern Han but rather high on Zhuang-Dai, indicating that they're mostly sinicized natives.

okarinaofsteiner
10-16-2021, 07:12 AM
A former Anthroscape member sent me a chart of 23mofang averages for various suburban and rural districts in China, which are believed to be more autosomally "representative" of specific regions and linguistic subgroups (as opposed to the city centers, which are more cosmopolitan and therefore more "mixed".) https://imgur.com/a/pxuhh3D

https://i.imgur.com/1UYN0Tn.jpg

Map (latitude + longitude) with ChinaMAP study clusters added
https://i.imgur.com/GX11i2q.png

I attempted to model the MDLP K23b results for these averages earlier this year by substituting the listed ancestry components with various combinations of MDLP K23b reference populations. For example, I would model "She" with MDLP K23b's She, "Tibetan" with the average of the 2 MDLP K23b "Tibetan" reference populations, etc.


N_Han = 46.17% S_EA, 33.55% T_A, 18.43% AN, based on "Han_North" (HGDP)
S_Han v1 = 4/10 Han_ (HGDP) + 2/10 Chinese_Dai + 1/10 Tujia + 1/10 She + 1/10 Korean (avg) + 1/10 Vietnamese
S_Han v2 = 4/14 Han_ (HGDP) + 2/14 Korean (avg) + 2/14 Vietnamese + 2/14 Chinese_Dai + 1/14 Tujia + 1/14 She + 1/14 Ami + 1/14 Naxi

The N_Han reference component was modeled as something slightly more "NE Asian" shifted than most actual Northern Han, since this was highest among the Shandong, Hebei, and southern Shanxi averages. The S_Han reference component was modeled as something in-between "Chinese_Taiwan"/"Hakka" and "Cantonese", since this component was highest among the Teochew, Leizhou, and Hakka-speaking area averages in Guangdong [the Yue-speaking areas scored lower on this component].

I decided to make another model for S_Han after seeing how my averages plotted on the Austronesian vs Tungus_Altaic graph against the Chinese samples from my original GEDmatch dataset. S_Han v2 matched the distribution better than S_Han v1.



Austronesian vs Tungus_Altaic with original GEDmatch dataset (S_Han v1)
https://i.imgur.com/w1LLi62l.png

Austronesian vs Tungus_Altaic with original GEDmatch dataset (S_Han v2)
https://i.imgur.com/SE1mMcfl.png

In the new version, the Foshan and Jiangmen averages score quite closely to the "Cantonese" reference population on Tungus_Altaic and Austronesian. The Quanzhou average is fairly close to "Chinese_Taiwan", which I believe corresponds to native Minnan/Hokkien speakers from Taipei and should also be a good proxy for Minnan/Hokkien speakers in Fujian itself. The relative positions of the 23mofang location averages are the same in both the old and new versions.


N-S cline with original GEDmatch dataset + DNAConnect.org adoptees (S_Han v1)
https://i.imgur.com/YBwj794l.png

N-S cline with original GEDmatch dataset + DNAConnect.org adoptees (S_Han v2)
https://i.imgur.com/FhIURm2l.png

The newer version also seems to fit the distribution of actual Chinese samples on the N-S cline + East Eurasian percentages graph better. Which makes sense because I used a greater number of source populations for S_Han v2 that aren't 100% East Eurasian. I thought this would better simulate how the "S_Han" component seems to include some of the Daic-like ancestry (but not necessarily "Dai") that the Lingnan/Min+Hakka-speaking Han groups which score highest on 23mofang's "Southern Han" seem to have.



https://i.imgur.com/Ejnw6QPl.png

https://i.imgur.com/APym4vql.png

Versions of the above with ChinaMAP clusters labeled. Gray = Southeast [Fujian], green = Hubei.

On the N-S cline graph, the Jianghuai Mandarin, Wu, and Mindong samples fall on the line between the Northern Han and Southern Han reference points. There also seems to be an inland-coastal cline in that the inland samples seem to be less pure East Eurasian; we can clearly see the transition from Min speakers (Fujian) to Gan speakers (Jiangxi), Xiang speakers (Hunan), and SW Mandarin speakers (Sichuan/Chongqing)- all of which seem similarly "southern" to some extent but as shown in the original 23mofang chart [in Chinese] have different affinities to different SEA-like ancestries.



https://i.imgur.com/ZJ0Z69E.png

Map showing where the 0.45 (blue), 0.50 (purple), and 0.55 (red) values of my N-S cline fall on my new model (S_Han v2). The 0.45 value corresponds almost exactly to the Qinling-Huaihe line. While the 0.55 value roughly corresponds to which Southern Han (rice area) groups are more SEA-influenced and which ones are more "Central"

Max_H
10-17-2021, 08:14 PM
When inland samples score less "East Eurasian" what component increases instead?

okarinaofsteiner
10-18-2021, 01:40 AM
When inland samples score less "East Eurasian" what component increases instead?

Looking at the original chart, the Southwest Mandarin, Xiang, and (to a lesser extent) Gan speaking averages generally score higher on "Hmong-Miao", "Tibeto-Burman", and "Lahu". The Sichuan/Chongqing/Guizhou averages also score higher on Daic (Tai-Kradai) than other Southwest Mandarin groups.

There is also a north-south gradient of increasing "Lahu" between Sichuan/Chongqing and everywhere to the south, as well as a northeast-southwest gradient between Gan, Xiang, and the previously mentioned Southwest Mandarin groups from Yunnan, Guizhou, etc. The Yue groups also score higher on "Lahu" than the Hakka groups, who score higher than the Gan groups, who average slightly higher than the Fujian Min groups + Shantou. "Lahu" is a Himalayan-speaking ethnic group, but it seems to be a proxy for Austroasiatic-like ancestry?

The Fujian Min samples score relatively high on "Korean", while also scoring lower than the surrounding Southern Han groups on almost every other non-Han component. The Hakka groups also have relatively high affinities with "Korean", although not as much as the Jianghuai Mandarin + Wu groups or the other Yellow Sea-area Northern Han groups.

This all means the inland Southern Han groups score slightly higher on MDLP K23b "Australoid" and "Melano-Polynesian", although given which reference populations were used to simulate these samples, some of the more "northern-shifted" Sichuanese averages also seem to have higher West Eurasian components too. The inland Northern Han groups score more West Eurasian (specifically Central Asian-like) components due to scoring higher on 23mofang "Mongol".

okarinaofsteiner
10-18-2021, 05:40 AM
When inland samples score less "East Eurasian" what component increases instead?

The thing about this model is that the input data is averages of various suburban/rural districts and counties in China. And I'm using averages of MDLP K23b reference populations to simulate the 23mofang ancestry components. So this doesn't predict how individuals will score very well.

https://i.imgur.com/APym4vql.png

Honestly the differences in "East Eurasian" (which is basically other people's definition of "East Asian") aren't that significant between the Sichuan averages and Fujian averages. 98.8% and 99.2%. The differences between the Northwestern Han averages and North China Plain Northern Han averages are much larger.

MNOPSC1b
10-19-2021, 12:02 AM
Looking at the original chart, the Southwest Mandarin, Xiang, and (to a lesser extent) Gan speaking averages generally score higher on "Hmong-Miao", "Tibeto-Burman", and "Lahu". The Sichuan/Chongqing/Guizhou averages also score higher on Daic (Tai-Kradai) than other Southwest Mandarin groups.

There is also a north-south gradient of increasing "Lahu" between Sichuan/Chongqing and everywhere to the south, as well as a northeast-southwest gradient between Gan, Xiang, and the previously mentioned Southwest Mandarin groups from Yunnan, Guizhou, etc. The Yue groups also score higher on "Lahu" than the Hakka groups, who score higher than the Gan groups, who average slightly higher than the Fujian Min groups + Shantou. "Lahu" is a Himalayan-speaking ethnic group, but it seems to be a proxy for Austroasiatic-like ancestry?

The Fujian Min samples score relatively high on "Korean", while also scoring lower than the surrounding Southern Han groups on almost every other non-Han component. The Hakka groups also have relatively high affinities with "Korean", although not as much as the Jianghuai Mandarin + Wu groups or the other Yellow Sea-area Northern Han groups.

This all means the inland Southern Han groups score slightly higher on MDLP K23b "Australoid" and "Melano-Polynesian", although given which reference populations were used to simulate these samples, some of the more "northern-shifted" Sichuanese averages also seem to have higher West Eurasian components too. The inland Northern Han groups score more West Eurasian (specifically Central Asian-like) components due to scoring higher on 23mofang "Mongol".

IMO, the Korean component most likely represents a eastern or coastal component, and it decreases as one moves further inland or further to the west.

It may also indicate that northern influence is stronger on the eastern and southeastern coast than the regions further to the west. I noticed years ago that the Northern Han component is significantly higher in Fujian and Chaoshan than in Western Guangdong and Guangxi. This might indicate that the historical migrations from Northern China to Southern China largely followed a coastal route, as opposed to moving across inland regions and the Nanling Range.

As for the Lahu component, I agree with you that it likely represents some sort of Austroasiatic influence. Despite speaking a Sino-Tibetan language, the Lahu people are quite southern-shifted on the autosomal graph, and their high frequencies of Y-chromosome haplogroup F2 also make them rather unique among Sino-Tibetan speaking peoples.

Max_H
10-20-2021, 09:04 PM
Looking at the original chart, the Southwest Mandarin, Xiang, and (to a lesser extent) Gan speaking averages generally score higher on "Hmong-Miao", "Tibeto-Burman", and "Lahu". The Sichuan/Chongqing/Guizhou averages also score higher on Daic (Tai-Kradai) than other Southwest Mandarin groups.

There is also a north-south gradient of increasing "Lahu" between Sichuan/Chongqing and everywhere to the south, as well as a northeast-southwest gradient between Gan, Xiang, and the previously mentioned Southwest Mandarin groups from Yunnan, Guizhou, etc. The Yue groups also score higher on "Lahu" than the Hakka groups, who score higher than the Gan groups, who average slightly higher than the Fujian Min groups + Shantou. "Lahu" is a Himalayan-speaking ethnic group, but it seems to be a proxy for Austroasiatic-like ancestry?

The Fujian Min samples score relatively high on "Korean", while also scoring lower than the surrounding Southern Han groups on almost every other non-Han component. The Hakka groups also have relatively high affinities with "Korean", although not as much as the Jianghuai Mandarin + Wu groups or the other Yellow Sea-area Northern Han groups.

This all means the inland Southern Han groups score slightly higher on MDLP K23b "Australoid" and "Melano-Polynesian", although given which reference populations were used to simulate these samples, some of the more "northern-shifted" Sichuanese averages also seem to have higher West Eurasian components too. The inland Northern Han groups score more West Eurasian (specifically Central Asian-like) components due to scoring higher on 23mofang "Mongol".

So in MDLP K23b East Eurasian is meant as East Asian? I know 23mofang but I haven't paid much attention to genetic testing kits and results yet... Lahu is probably Austroasiatic-related indeed as I said before I think there is a difference among inland and coastal Chinese groups in their southern component, inland scoring more Mekong_N-related and coastal scoring more Fujian_N-related.

I also think that Mekong_N-related components (a type of ancestry widespread in Austroasiatic groups) are a bit more Onge-like/southern-shifted compared to coastal Neolithic Fujian populations.

Korean seems like an eastern coastal component, perhaps also due to the southern ancestry in Koreans being more related to the Fujian-type of southern East Asian ancestry than to the Mekong-type.

I am not very sure about West Eurasian ancestry in Sichuanese samples (not to say that they don't have it, just usually don't show it) , except if they are Qiang that I've noticed sometimes score some low West Eurasian-related ancestry. But some Sichuanese have Tibetan-like ancestry so in this case, I can see how they would get excess of West Eurasian or Indian admixture relative to coastal southern Han.

okarinaofsteiner
10-20-2021, 10:47 PM
So in MDLP K23b East Eurasian is meant as East Asian? I know 23mofang but I haven't paid much attention to genetic testing kits and results yet... Lahu is probably Austroasiatic-related indeed as I said before I think there is a difference among inland and coastal Chinese groups in their southern component, inland scoring more Mekong_N-related and coastal scoring more Fujian_N-related.

I also think that Mekong_N-related components (a type of ancestry widespread in Austroasiatic groups) are a bit more Onge-like/southern-shifted compared to coastal Neolithic Fujian populations.

My "East Eurasian" for MDLP K23b is just "Austronesian", "South_East_Asian", "Tungus_Altaic", "East_Siberian", and "Paleo_Siberian". So basically just the "Mongoloid" ancestry components associated with East Asians proper- as opposed to any of the indigenous peoples of the Americas, or the paraphyletic "Australoid" groups [AASI proper, Onge, Hoabinhian, Malaysian Negrito, Philippine Negrito, Australian aborigine, Papuan, the diverged ancestry component of Polynesians, the diverged component in Tibetans, the diverged component in Jomon, etc]

I don't consider being more "Onge-like" to be more "southern-shifted", at least if we're talking about strictly "Basal East Asian" ancestry. We know the ancestors of Austroasiatic speakers mixed with Onge-like populations while still in modern-day China, but this probably took place after the inland-coastal split among southern East Asians that took place over 10k years ago.

I wouldn't be surprised if Mekong_N has some shared drift in its "Basal East Asian" with the ancestors of Himalayans/Tibeto-Burmans that Fujian_N doesn't have. But this is all just guesswork on my part. I haven't had time to read or go back to the papers that discuss this in detail.

I don't think any of the 23mofang reference components have particularly strong affinities with Fujian_N aside from maybe Daic (Tai-Kradai), which some posters have suggested is genetically transitional between "Proto-Austroasiatic" and "Proto-Austronesian". But I agree that whatever "southern" affinities Korean and Japanese have are going to be with Fujian_N related groups, not Mekong_N related groups.

MNOPSC1b
10-20-2021, 11:44 PM
So in MDLP K23b East Eurasian is meant as East Asian? I know 23mofang but I haven't paid much attention to genetic testing kits and results yet... Lahu is probably Austroasiatic-related indeed as I said before I think there is a difference among inland and coastal Chinese groups in their southern component, inland scoring more Mekong_N-related and coastal scoring more Fujian_N-related.

I also think that Mekong_N-related components (a type of ancestry widespread in Austroasiatic groups) are a bit more Onge-like/southern-shifted compared to coastal Neolithic Fujian populations.

Korean seems like an eastern coastal component, perhaps also due to the southern ancestry in Koreans being more related to the Fujian-type of southern East Asian ancestry than to the Mekong-type.

Largely agree, though I think the similarity between Koreans and Eastern Chinese isn't because of Koreans being more related to the Fujian-type of southern East Asian ancestry, but rather because people from Jiangsu, Zhejiang, and Fujian have received significant Northern Chinese / Neolithic Yellow River ancestry, which Koreans also have a fair amount. The actual Austronesian or Fujian_N related component among Koreans and Japanese is quite low, and Mekong_N related even lower. I don't think those are the causes of similarity.

Max_H
10-21-2021, 09:18 AM
Largely agree, though I think the similarity between Koreans and Eastern Chinese isn't because of Koreans being more related to the Fujian-type of southern East Asian ancestry, but rather because people from Jiangsu, Zhejiang, and Fujian have received significant Northern Chinese / Neolithic Yellow River ancestry, which Koreans also have a fair amount. The actual Austronesian or Fujian_N related component among Koreans and Japanese is quite low, and Mekong_N related even lower. I don't think those are the causes of similarity.

I agree with eastern Chinese having received significant Neolithic Yellow River ancestry, but NW Han such as Shanxi Han are even more northern-shifted yet do not score similar as similar to Koreans as eastern Han too (nor do they resemble them more phenotypically-but this is a different story). So I was wondering if the explanation is sharing of a southern component instead.

Not sure Koreans even score above-trace-level Mekong_N related components.


Edit: An interesting possibility is higher Boshan_N-related (coastal Neolithic northern East Asian ancestry) in eastern Han relative to inland Han (including northern inland Han) which is probably also found in substantial amounts in Koreans. At least this looks to be the case based on Global 25.

Max_H
10-21-2021, 09:23 AM
My "East Eurasian" for MDLP K23b is just "Austronesian", "South_East_Asian", "Tungus_Altaic", "East_Siberian", and "Paleo_Siberian". So basically just the "Mongoloid" ancestry components associated with East Asians proper- as opposed to any of the indigenous peoples of the Americas, or the paraphyletic "Australoid" groups [AASI proper, Onge, Hoabinhian, Malaysian Negrito, Philippine Negrito, Australian aborigine, Papuan, the diverged ancestry component of Polynesians, the diverged component in Tibetans, the diverged component in Jomon, etc]

I don't consider being more "Onge-like" to be more "southern-shifted", at least if we're talking about strictly "Basal East Asian" ancestry. We know the ancestors of Austroasiatic speakers mixed with Onge-like populations while still in modern-day China, but this probably took place after the inland-coastal split among southern East Asians that took place over 10k years ago.

I wouldn't be surprised if Mekong_N has some shared drift in its "Basal East Asian" with the ancestors of Himalayans/Tibeto-Burmans that Fujian_N doesn't have. But this is all just guesswork on my part. I haven't had time to read or go back to the papers that discuss this in detail.

I don't think any of the 23mofang reference components have particularly strong affinities with Fujian_N aside from maybe Daic (Tai-Kradai), which some posters have suggested is genetically transitional between "Proto-Austroasiatic" and "Proto-Austronesian". But I agree that whatever "southern" affinities Korean and Japanese have are going to be with Fujian_N related groups, not Mekong_N related groups.

What else could explain the southern pull of populations carrying Mekong_N-related ancestry compared to eastern ones carrying more Fujian_LN-related? Since you mentioned that affinities with "Southeast Asian" groups are included in MDLP K23b "East Eurasian" category.

23mofang components all look to reflect modern-day populations, but I think the Daic component is indeed transitional between "Proto-Austroasitic" and "Proto-Austronesian".

okarinaofsteiner
10-21-2021, 04:27 PM
What else could explain the southern pull of populations carrying Mekong_N-related ancestry compared to eastern ones carrying more Fujian_LN-related? Since you mentioned that affinities with "Southeast Asian" groups are included in MDLP K23b "East Eurasian" category.

As far as 23mofang is concerned, coastal Southern Han groups seem to have less SEA-like ancestry overall than inland Southern Han groups. Tbh none of the other "southern" components have particularly strong affinities with Fujian_N. I don't see why Hmong-Mien or Southern Han (which is highest in the coastal area between southern Fujian and the Leizhou Peninsula) would be better proxies for Fujian_N than Daic (Tai-Kradai).

The Fujian Min averages have slightly higher levels of Daic than the Wu averages (1.1-1.2% vs 0.9-1.0%), both of which are lower than the Gan averages, which are lower than the Xiang averages, which are around the same as the Hakka averages. It seems to me that the "southern" ancestry Fujian Han have is mostly in the "Southern Han" component, which isn't as good a fit for the "southern" ancestry in other Southern Han groups, so their "southern" ancestry tend to be modeled more as various "non-Han" components like Daic and Hmong-Mien.

MNOPSC1b
10-21-2021, 11:38 PM
I agree with eastern Chinese having received significant Neolithic Yellow River ancestry, but NW Han such as Shanxi Han are even more northern-shifted yet do not score similar as similar to Koreans as eastern Han too (nor do they resemble them more phenotypically-but this is a different story). So I was wondering if the explanation is sharing of a southern component instead.

Not sure Koreans even score above-trace-level Mekong_N related components.


Edit: An interesting possibility is higher Boshan_N-related (coastal Neolithic northern East Asian ancestry) in eastern Han relative to inland Han (including northern inland Han) which is probably also found in substantial amounts in Koreans. At least this looks to be the case based on Global 25.

The reason is most likely just as you mentioned, Koreans and Eastern Chinese inherited more coastal Neolithic Northern East Asian ancestry (represented by Boshan and Xiaojingshan). On the other hand, inland Northern Han inherited more Upper Yellow River Northern East Asian ancestry or steppe-related ancestry.

MNOPSC1b
10-21-2021, 11:56 PM
As far as 23mofang is concerned, coastal Southern Han groups seem to have less SEA-like ancestry overall than inland Southern Han groups. Tbh none of the other "southern" components have particularly strong affinities with Fujian_N. I don't see why Hmong-Mien or Southern Han (which is highest in the coastal area between southern Fujian and the Leizhou Peninsula) would be better proxies for Fujian_N than Daic (Tai-Kradai).

The Fujian Min averages have slightly higher levels of Daic than the Wu averages (1.1-1.2% vs 0.9-1.0%), both of which are lower than the Gan averages, which are lower than the Xiang averages, which are around the same as the Hakka averages. It seems to me that the "southern" ancestry Fujian Han have is mostly in the "Southern Han" component, which isn't as good a fit for the "southern" ancestry in other Southern Han groups, so their "southern" ancestry tend to be modeled more as various "non-Han" components like Daic and Hmong-Mien.

For the East-West difference among Southern Chinese I don't really see it as merely coastal vs inland but is more complicated than that. As I mentioned numerous times here and on Anthroscape, there's a clear decrease in Northern Han component and a clear increase in Daic component as we move from Fujian to Guangxi, with the highest concentration of Daic component found among the regional Cantonese populations in Western Guangdong and Guangxi, can reach to more than 20% according to some statistics I've seen, which is way higher than the Daic component found among Fujianese. And on average these regional Cantonese speakers are the southernmost shifted out of all Han Chinese groups, and are very close to SE Asians. IMO they're in fact sinicized Zhuang, and should really be granted minority status.

However, once we get to more inland regions like Guizhou and Yunnan, the Northern Han component starts to increase again as well as the Yi and Hmong-Mien components, but the Daic component decreases.

Max_H
10-27-2021, 03:06 PM
For the East-West difference among Southern Chinese I don't really see it as merely coastal vs inland but is more complicated than that. As I mentioned numerous times here and on Anthroscape, there's a clear decrease in Northern Han component and a clear increase in Daic component as we move from Fujian to Guangxi, with the highest concentration of Daic component found among the regional Cantonese populations in Western Guangdong and Guangxi, can reach to more than 20% according to some statistics I've seen, which is way higher than the Daic component found among Fujianese. And on average these regional Cantonese speakers are the southernmost shifted out of all Han Chinese groups, and are very close to SE Asians. IMO they're in fact sinicized Zhuang, and should really be granted minority status.

However, once we get to more inland regions like Guizhou and Yunnan, the Northern Han component starts to increase again as well as the Yi and Hmong-Mien components, but the Daic component decreases.

Do you think the west-east differentiation of southern Chinese is then in part at least due to different amounts of northern ancestry as well as coastal vs inland southern East Asian ancestry?

Max_H
10-27-2021, 03:08 PM
The reason is most likely just as you mentioned, Koreans and Eastern Chinese inherited more coastal Neolithic Northern East Asian ancestry (represented by Boshan and Xiaojingshan). On the other hand, inland Northern Han inherited more Upper Yellow River Northern East Asian ancestry or steppe-related ancestry.

I think the Steppe-related ancestry in modern-day northern Han is mostly a later phenomenon because it is very low to non-present in southern China. Shanxi Han are probably on average 4-5% West Eurasian but even in Shandong it is a lot lower.

okarinaofsteiner
10-28-2021, 05:23 AM
Do you think the west-east differentiation of southern Chinese is then in part at least due to different amounts of northern ancestry as well as coastal vs inland southern East Asian ancestry?

I think the differences between Fujian and the Pearl River Delta have more to do with different amounts of northern ancestry, but for Sichuan vs Yunnan vs Guizhou vs Hunan vs Jiangxi, it's likely different types of "southern" (non-Han) East Asian ancestries.


I think the Steppe-related ancestry in modern-day northern Han is mostly a later phenomenon because it is very low to non-present in southern China. Shanxi Han are probably on average 4-5% West Eurasian but even in Shandong it is a lot lower.

Yup, Razib Khan mentioned this 2 years ago (https://youtu.be/uR2puKcym60?t=995). That being said, I have noticed in my private GEDmatch samples and in the DNAConnect.org adoptees that Guangdong/Hong Kong samples have slightly higher levels of trace West Eurasian than Fujianese/Taiwanese. Guangdong and Pearl River Delta Han do seem to score slightly lower on East Eurasian than Fujianese and Taiwanese Han, but this could just as much be due to sample size issues as different levels of South Eurasian and West Eurasian trace ancestry.

From the WBBC study this year (https://anthrogenica.com/showthread.php?709-New-DNA-Papers/page158&p=798518#post798518)- smaller numbers/red = less genetic distance, larger numbers/blue = more genetic distance. Hunan, Jiangxi, Zhejiang, and Fujian being among the most "pure East Eurasian" provinces with the lowest amounts of non-East Asian affinities is consistent with what I've noticed in my private GEDmatch samples- the Han Chinese who score highest on East Eurasian are the Southern Han who are more southern than Yangtze Delta Han but more northern than Lingnan Han (https://anthrogenica.com/showthread.php?20657-okarinaofsteiner-s-East-Eurasian-GEDmatch-megathread&p=770762&viewfull=1#post770762).

https://i.imgur.com/NJabEYy.png

MNOPSC1b
10-28-2021, 12:45 PM
I think the differences between Fujian and the Pearl River Delta have more to do with different amounts of northern ancestry, but for Sichuan vs Yunnan vs Guizhou vs Hunan vs Jiangxi, it's likely different types of "southern" (non-Han) East Asian ancestries.

There're two vectors of difference between Fujianese and people of Guangdong and Guangxi. First is that Fujianese have significantly higher Northern Han ancestry than those from Pearl River Delta and Western Guangdong. Second is that the Tai-Kradai ancestry of Fujianese is significantly lower than that of Cantonese-speaking populations. So it isn't just a difference in northern ancestry, but also a difference in Kradai-related ancestry.

Max_H
10-30-2021, 10:36 PM
I think the differences between Fujian and the Pearl River Delta have more to do with different amounts of northern ancestry, but for Sichuan vs Yunnan vs Guizhou vs Hunan vs Jiangxi, it's likely different types of "southern" (non-Han) East Asian ancestries.



Yup, Razib Khan mentioned this 2 years ago (https://youtu.be/uR2puKcym60?t=995). That being said, I have noticed in my private GEDmatch samples and in the DNAConnect.org adoptees that Guangdong/Hong Kong samples have slightly higher levels of trace West Eurasian than Fujianese/Taiwanese. Guangdong and Pearl River Delta Han do seem to score slightly lower on East Eurasian than Fujianese and Taiwanese Han, but this could just as much be due to sample size issues as different levels of South Eurasian and West Eurasian trace ancestry.

From the WBBC study this year (https://anthrogenica.com/showthread.php?709-New-DNA-Papers/page158&p=798518#post798518)- smaller numbers/red = less genetic distance, larger numbers/blue = more genetic distance. Hunan, Jiangxi, Zhejiang, and Fujian being among the most "pure East Eurasian" provinces with the lowest amounts of non-East Asian affinities is consistent with what I've noticed in my private GEDmatch samples- the Han Chinese who score highest on East Eurasian are the Southern Han who are more southern than Yangtze Delta Han but more northern than Lingnan Han (https://anthrogenica.com/showthread.php?20657-okarinaofsteiner-s-East-Eurasian-GEDmatch-megathread&p=770762&viewfull=1#post770762).

https://i.imgur.com/NJabEYy.png

Agreed, but I think the lower Fst to West Eurasians or Africans is both a function of West Eurasian trace admixture and southern Han carrying more "basal" ancestry IMO. So Lingnan Han may not differ in their West Eurasian amounts to say Jiangxi Han

okarinaofsteiner
11-01-2021, 05:51 AM
https://i.imgur.com/NJabEYy.png
Agreed, but I think the lower Fst to West Eurasians or Africans is both a function of West Eurasian trace admixture and southern Han carrying more "basal" ancestry IMO. So Lingnan Han may not differ in their West Eurasian amounts to say Jiangxi Han

Guangdong has lower FST with Amerindian than all of its neighboring provinces, so you could argue that Guangdong Han are somewhat more Amerindian mixed thanks to post-Columbian Chinese migration to Latin American countries. For example, Chinese American YouTuber Bart Kwan has some Peruvian ancestry from a Guangdong ancestor who migrated to Peru (https://www.youtube.com/watch?v=hxyQFZzjSBI).


There're two vectors of difference between Fujianese and people of Guangdong and Guangxi. First is that Fujianese have significantly higher Northern Han ancestry than those from Pearl River Delta and Western Guangdong. Second is that the Tai-Kradai ancestry of Fujianese is significantly lower than that of Cantonese-speaking populations. So it isn't just a difference in northern ancestry, but also a difference in Kradai-related ancestry.

This makes sense if you accept that the SEA-like ancestry of Fujianese Han is Daic, which the presence of Daic origin common-use words in Minnan topolects (e.g. /sui⁵³/ = beautiful (https://en.wiktionary.org/wiki/%E5%AA%A0#Chinese), /baʔ³²/ = meat (https://en.wiktionary.org/wiki/%E8%82%89#Etymology_2)) would suggest.

Some Common Min words appear to have an Austroasiatic origin- the Min words for mango (https://en.wiktionary.org/wiki/%E6%AA%A8#Chinese) and child (https://en.wiktionary.org/wiki/%E5%9B%9D#Etymology_1) resemble their Vietnamese and Khmer equivalents.

Searching
11-01-2021, 05:53 PM
My mother’s Chinese match

HarappaWorld 4-Ancestors Oracle

This program is based on 4-Ancestors Oracle Version 0.96 by Alexandr Burnashev.
Questions about results should be sent to him at: [email protected]
Original concept proposed by Sergey Kozlov.
Many thanks to Alexandr for helping us get this web version developed.

23 April 2013 - Oracle reference population percentages revised.

Admix Results (sorted):

# Population Percent
1 NE-Asian 70.96
2 SE-Asian 13.96
3 Siberian 8.54
4 Caucasian 1.48
5 American 1.31
6 E-African 1.27
7 Baloch 1.10


Finished reading population data. 377 populations found.
16 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 naxi_hgdp @ 5.604744
2 yi_hgdp @ 6.029672
3 japanese_hgdp @ 6.145211
4 han-nchina_hgdp @ 7.871053
5 tu_hgdp @ 7.960814
6 chinese-beijing_1000genomes @ 8.973565
7 chinese-beijing_hapmap @ 9.224724
8 xibo_hgdp @ 15.481563
9 naga_metspalu @ 17.103523
10 tibet_simonson @ 17.210011
11 tujia_hgdp @ 17.559629
12 chinese-american_hapmap @ 18.134329
13 mongola_hgdp @ 18.509668
14 aonaga_reich @ 19.111969
15 nysha_reich @ 19.167425
16 han_hgdp @ 21.205057
17 han-chinese-south_1000genomes @ 21.941076
18 chinese_xing @ 22.382301
19 she_hgdp @ 23.454924
20 singapore-chinese_sgvp @ 25.318146

Using 2 populations approximation:
1 50% chinese-beijing_hapmap +50% xibo_hgdp @ 4.398166


Using 3 populations approximation:
1 50% han-nchina_hgdp +25% mongola_hgdp +25% tujia_hgdp @ 2.742718


Using 4 populations approximation:
++++++++++++++++++++++++++++++++++++++
1 chinese-american_hapmap + han-nchina_hgdp + tu_hgdp + xibo_hgdp @ 2.504567
2 han-nchina_hgdp + tu_hgdp + tujia_hgdp + xibo_hgdp @ 2.563042
3 han_hgdp + han-nchina_hgdp + tu_hgdp + xibo_hgdp @ 2.658762
4 han-chinese-south_1000genomes + han-nchina_hgdp + tu_hgdp + xibo_hgdp @ 2.678569
5 han-nchina_hgdp + han-nchina_hgdp + mongola_hgdp + tujia_hgdp @ 2.742718
6 chinese-american_hapmap + han-nchina_hgdp + han-nchina_hgdp + mongola_hgdp @ 2.748417
7 chinese-american_hapmap + han-nchina_hgdp + mongola_hgdp + tu_hgdp @ 2.797407
8 chinese_xing + han-nchina_hgdp + tu_hgdp + xibo_hgdp @ 2.839340
9 han-nchina_hgdp + she_hgdp + tu_hgdp + xibo_hgdp @ 2.840758
10 chinese-beijing_1000genomes + tu_hgdp + tujia_hgdp + xibo_hgdp @ 2.851992
11 chinese-beijing_hapmap + tu_hgdp + tujia_hgdp + xibo_hgdp @ 2.875846
12 chinese-american_hapmap + chinese-beijing_1000genomes + tu_hgdp + xibo_hgdp @ 2.877853
13 han-nchina_hgdp + mongola_hgdp + tu_hgdp + tujia_hgdp @ 2.887650
14 chinese-american_hapmap + chinese-beijing_hapmap + tu_hgdp + xibo_hgdp @ 2.916813
15 chinese-american_hapmap + han-nchina_hgdp + japanese_hgdp + xibo_hgdp @ 2.935339
16 chinese-beijing_hapmap + chinese-beijing_hapmap + mongola_hgdp + tu_hgdp @ 2.959428
17 han-nchina_hgdp + han-nchina_hgdp + tujia_hgdp + xibo_hgdp @ 2.959466
18 chinese-beijing_1000genomes + chinese-beijing_hapmap + mongola_hgdp + tu_hgdp @ 2.965104
19 aonaga_reich + han-nchina_hgdp + miao_hgdp + mongola_hgdp @ 2.968184
20 han-nchina_hgdp + japanese_hgdp + tujia_hgdp + xibo_hgdp @ 2.981574

I believe this lady is from Changzhou. I remember growing up hearing of a distant ancestor who was Chinese.

MNOPSC1b
11-01-2021, 11:40 PM
Guangdong has lower FST with Amerindian than all of its neighboring provinces, so you could argue that Guangdong Han are somewhat more Amerindian mixed thanks to post-Columbian Chinese migration to Latin American countries. For example, Chinese American YouTuber Bart Kwan has some Peruvian ancestry from a Guangdong ancestor who migrated to Peru (https://www.youtube.com/watch?v=hxyQFZzjSBI).



This makes sense if you accept that the SEA-like ancestry of Fujianese Han is Daic, which the presence of Daic origin common-use words in Minnan topolects (e.g. /sui⁵³/ = beautiful (https://en.wiktionary.org/wiki/%E5%AA%A0#Chinese), /baʔ³²/ = meat (https://en.wiktionary.org/wiki/%E8%82%89#Etymology_2)) would suggest.

Some Common Min words appear to have an Austroasiatic origin- the Min words for mango (https://en.wiktionary.org/wiki/%E6%AA%A8#Chinese) and child (https://en.wiktionary.org/wiki/%E5%9B%9D#Etymology_1) resemble their Vietnamese and Khmer equivalents.

I'm not sure if you were meant to be joking or serious when you said that Guangdong Han are more Amerindian-admixed compared to other Chinese. The origin of Amerindians had nothing to do with Cantonese, they originated from somewhere in Siberia and diverged from the ancestors of modern East Asians at least 27k years ago, and they carry around 30% ANE. Just like you said, modern-day Amerindians (especially those from Peru but also from other areas like Mexico or the west coast of US and CA) might be more Cantonese-admixed due to recent Cantonese migrations to the Americas.

I also feel that you have a tendency to forget what you've posted. In fact most of what I use to refute your arguments comes directly from what you've posted earlier. Here's a table about the admixtures in different Han sub-populations that you've posted in this thread.

https://i.imgur.com/1UYN0Tn.jpg

According to the picture Min-speaking peoples indeed have Tai-Kradai admixtures, however they carry them in very low amounts, merely around 1%. The only exception is the Zhanjiang Min speakers who carry 11.8% Daic admixtures, but judging by the fact that Zhanjiang is a city located on the Leizhou peninsula on the southwestern extreme of Guangdong bordering Guangxi and Hainan, such a result wouldn't be surprising.

For the Yue speakers however, even the one with the lowest Daic admixtures still carry 15.2% of them, whereas the one with the highest Daic admixtures reaches 26%. You can practically call them sinicized Dai or sinicized Zhuang with such high proportions of Daic admixtures. And the shared vocabulary between Yue and Daic languages are many, a lot more than the two you found for Min.

okarinaofsteiner
11-02-2021, 02:43 AM
My mother’s Chinese match

HarappaWorld 4-Ancestors Oracle

This program is based on 4-Ancestors Oracle Version 0.96 by Alexandr Burnashev.
Questions about results should be sent to him at: [email protected]
Original concept proposed by Sergey Kozlov.
Many thanks to Alexandr for helping us get this web version developed.

23 April 2013 - Oracle reference population percentages revised.

Admix Results (sorted):

# Population Percent
1 NE-Asian 70.96
2 SE-Asian 13.96
3 Siberian 8.54
4 Caucasian 1.48
5 American 1.31
6 E-African 1.27
7 Baloch 1.10


Finished reading population data. 377 populations found.
16 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 naxi_hgdp @ 5.604744
2 yi_hgdp @ 6.029672
3 japanese_hgdp @ 6.145211
4 han-nchina_hgdp @ 7.871053
5 tu_hgdp @ 7.960814
6 chinese-beijing_1000genomes @ 8.973565
7 chinese-beijing_hapmap @ 9.224724
8 xibo_hgdp @ 15.481563
9 naga_metspalu @ 17.103523
10 tibet_simonson @ 17.210011
11 tujia_hgdp @ 17.559629
12 chinese-american_hapmap @ 18.134329
13 mongola_hgdp @ 18.509668
14 aonaga_reich @ 19.111969
15 nysha_reich @ 19.167425
16 han_hgdp @ 21.205057
17 han-chinese-south_1000genomes @ 21.941076
18 chinese_xing @ 22.382301
19 she_hgdp @ 23.454924
20 singapore-chinese_sgvp @ 25.318146

Using 2 populations approximation:
1 50% chinese-beijing_hapmap +50% xibo_hgdp @ 4.398166


Using 3 populations approximation:
1 50% han-nchina_hgdp +25% mongola_hgdp +25% tujia_hgdp @ 2.742718


Using 4 populations approximation:
++++++++++++++++++++++++++++++++++++++
1 chinese-american_hapmap + han-nchina_hgdp + tu_hgdp + xibo_hgdp @ 2.504567
2 han-nchina_hgdp + tu_hgdp + tujia_hgdp + xibo_hgdp @ 2.563042
3 han_hgdp + han-nchina_hgdp + tu_hgdp + xibo_hgdp @ 2.658762
4 han-chinese-south_1000genomes + han-nchina_hgdp + tu_hgdp + xibo_hgdp @ 2.678569
5 han-nchina_hgdp + han-nchina_hgdp + mongola_hgdp + tujia_hgdp @ 2.742718
6 chinese-american_hapmap + han-nchina_hgdp + han-nchina_hgdp + mongola_hgdp @ 2.748417
7 chinese-american_hapmap + han-nchina_hgdp + mongola_hgdp + tu_hgdp @ 2.797407
8 chinese_xing + han-nchina_hgdp + tu_hgdp + xibo_hgdp @ 2.839340
9 han-nchina_hgdp + she_hgdp + tu_hgdp + xibo_hgdp @ 2.840758
10 chinese-beijing_1000genomes + tu_hgdp + tujia_hgdp + xibo_hgdp @ 2.851992
11 chinese-beijing_hapmap + tu_hgdp + tujia_hgdp + xibo_hgdp @ 2.875846
12 chinese-american_hapmap + chinese-beijing_1000genomes + tu_hgdp + xibo_hgdp @ 2.877853
13 han-nchina_hgdp + mongola_hgdp + tu_hgdp + tujia_hgdp @ 2.887650
14 chinese-american_hapmap + chinese-beijing_hapmap + tu_hgdp + xibo_hgdp @ 2.916813
15 chinese-american_hapmap + han-nchina_hgdp + japanese_hgdp + xibo_hgdp @ 2.935339
16 chinese-beijing_hapmap + chinese-beijing_hapmap + mongola_hgdp + tu_hgdp @ 2.959428
17 han-nchina_hgdp + han-nchina_hgdp + tujia_hgdp + xibo_hgdp @ 2.959466
18 chinese-beijing_1000genomes + chinese-beijing_hapmap + mongola_hgdp + tu_hgdp @ 2.965104
19 aonaga_reich + han-nchina_hgdp + miao_hgdp + mongola_hgdp @ 2.968184
20 han-nchina_hgdp + japanese_hgdp + tujia_hgdp + xibo_hgdp @ 2.981574

I believe this lady is from Changzhou. I remember growing up hearing of a distant ancestor who was Chinese.

Very interesting HarappaWorld results! Her Siberian score is pretty high for someone who is allegedly from the Yangtze Delta area. The range for Northern Han (as in north of the Qinling-Huaihe line) seems to be 6-10%.

Do you have her MDLP K23b results by any chance?

okarinaofsteiner
11-02-2021, 02:53 AM
I'm not sure if you were meant to be joking or serious when you said that Guangdong Han are more Amerindian-admixed compared to other Chinese. The origin of Amerindians had nothing to do with Cantonese, they originated from somewhere in Siberia and diverged from the ancestors of modern East Asians at least 27k years ago, and they carry around 30% ANE. Just like you said, modern-day Amerindians (especially those from Peru but also from other areas like Mexico or the west coast of US and CA) might be more Cantonese-admixed due to recent Cantonese migrations to the Americas.

Modern-day Amerindians who are more culturally isolated definitely are not Chinese/Filipino/Japanese mixed lmao. The culturally dominant mestizos (mixed-race population) are a different story.

I'm completely serious though, the Guangdong FST with AMR on that subgraph is a greener color (less blue) than Guangxi, Hunan, Jiangxi, and Fujian. Although Guangdong is still higher (more blue) than Henan or Hebei, which are clearly light-green on the map.

MNOPSC1b
11-02-2021, 11:04 PM
Modern-day Amerindians who are more culturally isolated definitely are not Chinese/Filipino/Japanese mixed lmao. The culturally dominant mestizos (mixed-race population) are a different story.

I'm completely serious though, the Guangdong FST with AMR on that subgraph is a greener color (less blue) than Guangxi, Hunan, Jiangxi, and Fujian. Although Guangdong is still higher (more blue) than Henan or Hebei, which are clearly light-green on the map.

Minor differences. GD still has a quite dark bluish green color similar to neighboring southern provinces, meaning they're still quite distant from Amerindians, unlike Central and Northern China who are shades lighter. The reason why GD is a tiny bit lighter than neighboring provinces might be due to the fact that the metropolis of the Pearl Delta have received some recent immigrants from Central and Northern China, particularly Shenzhen which has been transformed from a Canto/Hakka speaking small town to a pred Mandarin speaking metropolis due to recent migrations.

No matter how hard you try to associate Cantonese with non-local populations like Northern Chinese or even Amerindians, you cannot ignore the high proportions of native Daic components in their genome.

Searching
11-03-2021, 02:08 PM
Very interesting HarappaWorld results! Her Siberian score is pretty high for someone who is allegedly from the Yangtze Delta area. The range for Northern Han (as in north of the Qinling-Huaihe line) seems to be 6-10%.

Do you have her MDLP K23b results by any chance?

Yeah here are the results. Growing up I assumed we had distant Hakka ancestry. I was told they are the ones who came to Caribbean.

Admix Results (sorted):

# Population Percent
1 South_East_Asian 44.41
2 Tungus-Altaic 31.61
3 Austronesian 14.91
4 East_Siberian 4.12
5 Amerindian 1.86
6 East_African 1.27


Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Korean_KR_ @ 8.879153
2 Han_North_ @ 8.980083
3 Tu_ @ 9.457696
4 Han-Mandarin_ @ 11.460402
5 Korean_ @ 11.521450
6 Japanese_ML_ @ 14.496802
7 Ryukyuan_ @ 17.150736
8 Xibo_ @ 17.985901
9 Mongola_China_ @ 19.069126
10 Jinuo_ @ 21.320864
11 Japanese_ @ 21.511366
12 Naxi_ @ 21.607758
13 Hakka_ @ 21.761303
14 Tujia_ @ 21.831999
15 Yi_ @ 22.328150
16 Chinese_Taiwan_ @ 22.463327
17 Tibetian_Madou_ @ 22.476599
18 Paluang_ @ 22.669994
19 Tibetian_TTR_ @ 23.186916
20 Han_Singapore_ @ 23.295523

Using 2 populations approximation:
1 50% Korean_KR_ +50% Tu_ @ 4.017998


Using 3 populations approximation:
1 50% Han_North_ +25% Jinuo_ +25% Mongol_Khalkha_ @ 2.769178


Using 4 populations approximation:
++++++++++++++++++++++++++++++++++++++++++++++++++ +++++++++++++++++++++++++++++
1 Hezhen_ + She_ + Tu_ + Tu_ @ 2.338829
2 Daur_ + She_ + Tu_ + Tu_ @ 2.359536
3 Han-Mandarin_ + Korean_KR_ + Mongola_China_ + Yi_ @ 2.376791
4 Daur_ + Han_ + Han_North_ + Tu_ @ 2.383345
5 Hmong_ + Japanese_ML_ + Mongol_Khalkha_ + Naga_ @ 2.392039
6 Aonaga_ + Cantonese_ + Japanese_ML_ + Mongol_Khalkha_ @ 2.408614
7 Han_ + Mongola_China_ + Naxi_ + Xibo_ @ 2.414393
8 Daur_ + Hmong_Miao_ + Tu_ + Tu_ @ 2.416025
9 Hezhen_ + Hmong_Miao_ + Tu_ + Tu_ @ 2.417664
10 Han-Mandarin_ + Korean_KR_ + Mongola_China_ + Naxi_ @ 2.419817
11 Cantonese_ + Korean_KR_ + Mongol_Khalkha_ + Naga_ @ 2.421076
12 Han_ + Mongola_China_ + Xibo_ + Yi_ @ 2.445156
13 Han_ + Han_North_ + Mongol_Khalkha_ + Tu_ @ 2.445260
14 Han_Singapore_ + Korean_KR_ + Mongola_China_ + Tibetian_Madou_ @ 2.445333
15 Han_ + Hezhen_ + Tu_ + Tu_ @ 2.445534
16 Japanese_ML_ + Mongol_Khalkha_ + Tujia_ + Yi_ @ 2.445703
17 Han_Singapore_ + Korean_KR_ + Mongola_China_ + Tibetian_TTR_ @ 2.470982
18 Daur_ + Naga_ + Vietnamese_north_ + Xibo_ @ 2.475363
19 Han-Mandarin_ + Han_North_ + Mongola_China_ + Tu_ @ 2.490290
20 Aonaga_ + Daur_ + Vietnamese_ + Xibo_ @ 2.493130

Done.

okarinaofsteiner
11-04-2021, 01:43 AM
Yeah here are the results. Growing up I assumed we had distant Hakka ancestry. I was told they are the ones who came to Caribbean.

Admix Results (sorted):

# Population Percent
1 South_East_Asian 44.41
2 Tungus-Altaic 31.61
3 Austronesian 14.91
4 East_Siberian 4.12
5 Amerindian 1.86
6 East_African 1.27


Finished reading population data. 620 populations found.

Hmm, i’ve never seen any confirmed Han Chinese samples score that low on Austronesian… the average for southern Jiangsu (idk if there are any other Changzhou’s in China) would be somewhere between 22-26%. Maybe this person is ethnic Hui (Chinese-speaking Muslims, they usually have a little Central Asian ancestry)?

MNOPSC1b
11-04-2021, 02:15 AM
Yeah here are the results. Growing up I assumed we had distant Hakka ancestry. I was told they are the ones who came to Caribbean.

Admix Results (sorted):

# Population Percent
1 South_East_Asian 44.41
2 Tungus-Altaic 31.61
3 Austronesian 14.91
4 East_Siberian 4.12
5 Amerindian 1.86
6 East_African 1.27


Too northern for Hakka. Likely from Eastern or Northern China.

Shuzam87
11-04-2021, 04:43 AM
Hmm, I’ve never seen any confirmed Han Chinese samples score that low on Austronesian… the average for southern Jiangsu (idk if there are any other Changzhou’s in China) would be somewhere between 22-26%. Maybe this person is ethnic Hui (Chinese-speaking Muslims, they usually have a little Central Asian ancestry)?

Probably not. He does not seem to score any distinct West Eurasian ancestry.

Searching
11-13-2021, 08:32 AM
Here is one of my Chinese matches. I believe he is from Guangdong

K13 Oracle ref data revised 21 Nov 2013

Admix Results (sorted):

# Population Percent
1 East_Asian 73.60
2 Siberian 22.80
3 Red_Sea 1.76
4 Amerindian 1.54


Finished reading population data. 204 populations found.
13 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Yizu @ 5.213615
2 Naxi @ 6.348151
3 Tujia @ 9.665417
4 Japanese @ 11.923898
5 Miaozu @ 12.293698
6 Tu @ 13.794580
7 She @ 14.183617
8 Lahu @ 16.150291
9 Vietnamese @ 22.527840
10 Tibeto-Burman_Burmese @ 24.390671
11 Xibo @ 24.685841
12 Hezhen @ 26.086847
13 Cambodian @ 26.912516
14 Dai @ 27.458467
15 Malay @ 28.698317
16 Mongolian @ 50.608688
17 Kirgiz @ 54.068760
18 Uygur @ 55.417290
19 Kazakh @ 56.161499
20 Hazara @ 59.340611

Using 2 populations approximation:
1 50% Japanese +50% Miaozu @ 2.198127


Using 3 populations approximation:
1 50% Japanese +25% Miaozu +25% Miaozu @ 2.198127


Using 4 populations approximation:
++++++++++++++++++++++++++++++
1 Japanese + She + Tu + Tujia @ 2.164868
2 Hezhen + She + She + Yizu @ 2.186690
3 Japanese + Japanese + Miaozu + Miaozu @ 2.198127
4 Japanese + Miaozu + She + Tu @ 2.249152
5 Hezhen + Tujia + Tujia + Tujia @ 2.249335
6 Japanese + Japanese + Miaozu + She @ 2.295932
7 Japanese + Japanese + She + Tujia @ 2.304203
8 Japanese + Japanese + Miaozu + Tujia @ 2.309614
9 Japanese + She + She + Tu @ 2.330396
10 Dai + Hezhen + Japanese + Tujia @ 2.334177
11 She + She + Xibo + Yizu @ 2.339176
12 Japanese + Miaozu + Miaozu + Tu @ 2.342459
13 Hezhen + Miaozu + She + Yizu @ 2.357302
14 Hezhen + Naxi + She + She @ 2.360351
15 Dai + Hezhen + Japanese + She @ 2.381162
16 Japanese + Miaozu + Tu + Tujia @ 2.386417
17 Dai + Hezhen + Japanese + Miaozu @ 2.405756
18 Japanese + Japanese + Lahu + Tujia @ 2.408715
19 Hezhen + Miaozu + Tujia + Tujia @ 2.429794
20 Miaozu + She + Xibo + Yizu @ 2.443003

Done.

Elapsed time 1.0078 seconds.

MDLP K23b Oracle Rev 2014 Sep 16

Admix Results (sorted):

# Population Percent
1 South_East_Asian 46.54
2 Tungus-Altaic 29.36
3 Austronesian 20.10
4 Melano_Polynesian 1.62
5 South_Central_Asian 1.34


Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Han-Mandarin_ @ 6.632914
2 Han_North_ @ 7.564625
3 Korean_KR_ @ 10.826612
4 Tu_ @ 12.620770
5 Korean_ @ 13.903770
6 Tujia_ @ 16.000553
7 Hakka_ @ 16.032185
8 Jinuo_ @ 16.311775
9 Chinese_Taiwan_ @ 16.817802
10 Japanese_ML_ @ 17.186871
11 Han_Singapore_ @ 17.671244
12 Paluang_ @ 18.372627
13 Han_ @ 18.600760
14 Ryukyuan_ @ 19.712696
15 Naxi_ @ 20.669102
16 Yi_ @ 21.101274
17 Lawa_ @ 21.451939
18 She_ @ 22.330212
19 Miao_ @ 22.660694
20 Karen_ @ 22.681860

Using 2 populations approximation:
1 50% Jinuo_ +50% Korean_ @ 2.865271


Using 3 populations approximation:
1 50% Korean_ +25% Tibetian_TTR_ +25% Zhuang_ @ 1.941959


Using 4 populations approximation:
++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++++++
1 Han_North_ + Japanese_ + Naga_ + Tagalog_ @ 1.797468
2 Aonaga_ + Korean_ + Korean_ + Tagalog_ @ 1.816949
3 Jiamao_ + Korean_ + Korean_ + Tibetian_TTR_ @ 1.836923
4 Japanese_ + Japanese_ML_ + Naga_ + Vietnamese_central_ @ 1.851762
5 Cantonese_ + Han-Mandarin_ + Japanese_ + Tibetian_TTR_ @ 1.858635
6 Japanese_ + Korean_KR_ + Naga_ + Yong_ @ 1.883814
7 Japanese_ML_ + Jiamao_ + Korean_ + Naga_ @ 1.913381
8 Japanese_ + Japanese_ML_ + Nysha_ + Vietnamese_central_ @ 1.922335
9 Japanese_ + Naga_ + Tagalog_ + Tu_ @ 1.926772
10 Japanese_ + Korean_ + Tibetian_TTR_ + Vietnamese_central_ @ 1.929158
11 Han-Mandarin_ + Japanese_ + Miao_ + Tu_ @ 1.936087
12 Korean_ + Korean_ + Tibetian_TTR_ + Zhuang_ @ 1.941959
13 Aonaga_ + Japanese_ + Japanese_ML_ + Vietnamese_central_ @ 1.943330
14 Han_North_ + Hmong_ + Japanese_ + Tu_ @ 1.945039
15 Cantonese_ + Han_North_ + Korean_ + Tu_ @ 1.948241
16 Hakka_ + Hakka_ + Japanese_ + Tibetian_TTR_ @ 1.959553
17 Aonaga_ + Japanese_ + Jiamao_ + Korean_KR_ @ 1.981708
18 Jiamao_ + Korean_ + Korean_KR_ + Tibetian_TTR_ @ 1.987086
19 Han-Mandarin_ + Japanese_ + Tibetian_TTR_ + Yao_ @ 1.996159
20 Han_ + Han-Mandarin_ + Korean_ + Tu_ @ 1.999900

Done.

okarinaofsteiner
11-14-2021, 08:10 AM
Here is one of my Chinese matches. I believe he is from Guangdong

MDLP K23b Oracle Rev 2014 Sep 16

Admix Results (sorted):

# Population Percent
1 South_East_Asian 46.54
2 Tungus-Altaic 29.36
3 Austronesian 20.10
4 Melano_Polynesian 1.62
5 South_Central_Asian 1.34


Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Han-Mandarin_ @ 6.632914
2 Han_North_ @ 7.564625
3 Korean_KR_ @ 10.826612
4 Tu_ @ 12.620770
5 Korean_ @ 13.903770
6 Tujia_ @ 16.000553
7 Hakka_ @ 16.032185
8 Jinuo_ @ 16.311775
9 Chinese_Taiwan_ @ 16.817802
10 Japanese_ML_ @ 17.186871
11 Han_Singapore_ @ 17.671244
12 Paluang_ @ 18.372627
13 Han_ @ 18.600760
14 Ryukyuan_ @ 19.712696
15 Naxi_ @ 20.669102
16 Yi_ @ 21.101274
17 Lawa_ @ 21.451939
18 She_ @ 22.330212
19 Miao_ @ 22.660694
20 Karen_ @ 22.681860

Using 2 populations approximation:
1 50% Jinuo_ +50% Korean_ @ 2.865271


Using 3 populations approximation:
1 50% Korean_ +25% Tibetian_TTR_ +25% Zhuang_ @ 1.941959


Using 4 populations approximation:
++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++++++
1 Han_North_ + Japanese_ + Naga_ + Tagalog_ @ 1.797468
2 Aonaga_ + Korean_ + Korean_ + Tagalog_ @ 1.816949
3 Jiamao_ + Korean_ + Korean_ + Tibetian_TTR_ @ 1.836923
4 Japanese_ + Japanese_ML_ + Naga_ + Vietnamese_central_ @ 1.851762
5 Cantonese_ + Han-Mandarin_ + Japanese_ + Tibetian_TTR_ @ 1.858635
6 Japanese_ + Korean_KR_ + Naga_ + Yong_ @ 1.883814
7 Japanese_ML_ + Jiamao_ + Korean_ + Naga_ @ 1.913381
8 Japanese_ + Japanese_ML_ + Nysha_ + Vietnamese_central_ @ 1.922335
9 Japanese_ + Naga_ + Tagalog_ + Tu_ @ 1.926772
10 Japanese_ + Korean_ + Tibetian_TTR_ + Vietnamese_central_ @ 1.929158
11 Han-Mandarin_ + Japanese_ + Miao_ + Tu_ @ 1.936087
12 Korean_ + Korean_ + Tibetian_TTR_ + Zhuang_ @ 1.941959
13 Aonaga_ + Japanese_ + Japanese_ML_ + Vietnamese_central_ @ 1.943330
14 Han_North_ + Hmong_ + Japanese_ + Tu_ @ 1.945039
15 Cantonese_ + Han_North_ + Korean_ + Tu_ @ 1.948241
16 Hakka_ + Hakka_ + Japanese_ + Tibetian_TTR_ @ 1.959553
17 Aonaga_ + Japanese_ + Jiamao_ + Korean_KR_ @ 1.981708
18 Jiamao_ + Korean_ + Korean_KR_ + Tibetian_TTR_ @ 1.987086
19 Han-Mandarin_ + Japanese_ + Tibetian_TTR_ + Yao_ @ 1.996159
20 Han_ + Han-Mandarin_ + Korean_ + Tu_ @ 1.999900

Done.

That is a lot more NEA-shifted than most Guangdong Han- this person scores lower on Austronesian than I do! Out of ~80 Guangdong adoptees in my DNAConnect.org adoptee dataset, maybe 2 of them were northern shifted to the point where they scored equally high on Tungus_Altaic as Austronesian, if not higher.

MNOPSC1b
11-14-2021, 04:28 PM
Here is one of my Chinese matches. I believe he is from Guangdong

K13 Oracle ref data revised 21 Nov 2013

Admix Results (sorted):

# Population Percent
1 East_Asian 73.60
2 Siberian 22.80
3 Red_Sea 1.76
4 Amerindian 1.54


Finished reading population data. 204 populations found.
13 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Yizu @ 5.213615
2 Naxi @ 6.348151
3 Tujia @ 9.665417
4 Japanese @ 11.923898
5 Miaozu @ 12.293698
6 Tu @ 13.794580
7 She @ 14.183617
8 Lahu @ 16.150291
9 Vietnamese @ 22.527840
10 Tibeto-Burman_Burmese @ 24.390671
11 Xibo @ 24.685841
12 Hezhen @ 26.086847
13 Cambodian @ 26.912516
14 Dai @ 27.458467
15 Malay @ 28.698317
16 Mongolian @ 50.608688
17 Kirgiz @ 54.068760
18 Uygur @ 55.417290
19 Kazakh @ 56.161499
20 Hazara @ 59.340611

Using 2 populations approximation:
1 50% Japanese +50% Miaozu @ 2.198127


Using 3 populations approximation:
1 50% Japanese +25% Miaozu +25% Miaozu @ 2.198127


Using 4 populations approximation:
++++++++++++++++++++++++++++++
1 Japanese + She + Tu + Tujia @ 2.164868
2 Hezhen + She + She + Yizu @ 2.186690
3 Japanese + Japanese + Miaozu + Miaozu @ 2.198127
4 Japanese + Miaozu + She + Tu @ 2.249152
5 Hezhen + Tujia + Tujia + Tujia @ 2.249335
6 Japanese + Japanese + Miaozu + She @ 2.295932
7 Japanese + Japanese + She + Tujia @ 2.304203
8 Japanese + Japanese + Miaozu + Tujia @ 2.309614
9 Japanese + She + She + Tu @ 2.330396
10 Dai + Hezhen + Japanese + Tujia @ 2.334177
11 She + She + Xibo + Yizu @ 2.339176
12 Japanese + Miaozu + Miaozu + Tu @ 2.342459
13 Hezhen + Miaozu + She + Yizu @ 2.357302
14 Hezhen + Naxi + She + She @ 2.360351
15 Dai + Hezhen + Japanese + She @ 2.381162
16 Japanese + Miaozu + Tu + Tujia @ 2.386417
17 Dai + Hezhen + Japanese + Miaozu @ 2.405756
18 Japanese + Japanese + Lahu + Tujia @ 2.408715
19 Hezhen + Miaozu + Tujia + Tujia @ 2.429794
20 Miaozu + She + Xibo + Yizu @ 2.443003

Done.

Elapsed time 1.0078 seconds.

MDLP K23b Oracle Rev 2014 Sep 16

Admix Results (sorted):

# Population Percent
1 South_East_Asian 46.54
2 Tungus-Altaic 29.36
3 Austronesian 20.10
4 Melano_Polynesian 1.62
5 South_Central_Asian 1.34


Finished reading population data. 620 populations found.
23 components mode.

--------------------------------

Least-squares method.

Using 1 population approximation:
1 Han-Mandarin_ @ 6.632914
2 Han_North_ @ 7.564625
3 Korean_KR_ @ 10.826612
4 Tu_ @ 12.620770
5 Korean_ @ 13.903770
6 Tujia_ @ 16.000553
7 Hakka_ @ 16.032185
8 Jinuo_ @ 16.311775
9 Chinese_Taiwan_ @ 16.817802
10 Japanese_ML_ @ 17.186871
11 Han_Singapore_ @ 17.671244
12 Paluang_ @ 18.372627
13 Han_ @ 18.600760
14 Ryukyuan_ @ 19.712696
15 Naxi_ @ 20.669102
16 Yi_ @ 21.101274
17 Lawa_ @ 21.451939
18 She_ @ 22.330212
19 Miao_ @ 22.660694
20 Karen_ @ 22.681860

Using 2 populations approximation:
1 50% Jinuo_ +50% Korean_ @ 2.865271


Using 3 populations approximation:
1 50% Korean_ +25% Tibetian_TTR_ +25% Zhuang_ @ 1.941959


Using 4 populations approximation:
++++++++++++++++++++++++++++++++++++++++++++++++++ ++++++++++++++++++++++++++
1 Han_North_ + Japanese_ + Naga_ + Tagalog_ @ 1.797468
2 Aonaga_ + Korean_ + Korean_ + Tagalog_ @ 1.816949
3 Jiamao_ + Korean_ + Korean_ + Tibetian_TTR_ @ 1.836923
4 Japanese_ + Japanese_ML_ + Naga_ + Vietnamese_central_ @ 1.851762
5 Cantonese_ + Han-Mandarin_ + Japanese_ + Tibetian_TTR_ @ 1.858635
6 Japanese_ + Korean_KR_ + Naga_ + Yong_ @ 1.883814
7 Japanese_ML_ + Jiamao_ + Korean_ + Naga_ @ 1.913381
8 Japanese_ + Japanese_ML_ + Nysha_ + Vietnamese_central_ @ 1.922335
9 Japanese_ + Naga_ + Tagalog_ + Tu_ @ 1.926772
10 Japanese_ + Korean_ + Tibetian_TTR_ + Vietnamese_central_ @ 1.929158
11 Han-Mandarin_ + Japanese_ + Miao_ + Tu_ @ 1.936087
12 Korean_ + Korean_ + Tibetian_TTR_ + Zhuang_ @ 1.941959
13 Aonaga_ + Japanese_ + Japanese_ML_ + Vietnamese_central_ @ 1.943330
14 Han_North_ + Hmong_ + Japanese_ + Tu_ @ 1.945039
15 Cantonese_ + Han_North_ + Korean_ + Tu_ @ 1.948241
16 Hakka_ + Hakka_ + Japanese_ + Tibetian_TTR_ @ 1.959553
17 Aonaga_ + Japanese_ + Jiamao_ + Korean_KR_ @ 1.981708
18 Jiamao_ + Korean_ + Korean_KR_ + Tibetian_TTR_ @ 1.987086
19 Han-Mandarin_ + Japanese_ + Tibetian_TTR_ + Yao_ @ 1.996159
20 Han_ + Han-Mandarin_ + Korean_ + Tu_ @ 1.999900

Done.

The results are too northern for someone from Guangdong, you probably misidentified his origin.

And if this person is indeed from Guangdong, he's most likely to be someone from the Chaoshan (Teochew) or the Hakka region, but not from the Yue or Canto region. Chaoshan, Hakka, and Canto are three distinct ethnolinguistic groups inhabiting Guangdong, with Canto being the most Daic-shifted/the most native-like, Chaoshan (Teochew) being the most northern-shifted/the most Sinitic-like, and Hakka lies somewhere in between.

okarinaofsteiner
11-14-2021, 05:22 PM
The results are too northern for someone from Guangdong, you probably misidentified his origin.

And if this person is indeed from Guangdong, he's most likely to be someone from the Chaoshan (Teochew) or the Hakka region, but not from the Yue or Canto region. Chaoshan, Hakka, and Canto are three distinct ethnolinguistic groups inhabiting Guangdong, with Canto being the most Daic-shifted/the most native-like, Chaoshan (Teochew) being the most northern-shifted/the most Sinitic-like, and Hakka lies somewhere in between.

That person is way too northern-shifted for the Teochew and Hakka range. If they’re from Guangdong, they almost certainly have to be from the Pearl River Delta. That’s the only area of Guangdong that’s cosmopolitan enough to have a nonzero probability of finding someone who scores within the range of Northern Han.

MNOPSC1b
11-14-2021, 09:45 PM
That person is way too northern-shifted for the Teochew and Hakka range. If they’re from Guangdong, they almost certainly have to be from the Pearl River Delta. That’s the only area of Guangdong that’s cosmopolitan enough to have a nonzero probability of finding someone who scores within the range of Northern Han.

Yep, agreed with you this time. The person in question is most likely a recent northern migrant living in the Pearl River Delta.

MNOPSC1b
11-25-2021, 12:59 AM
Very interesting contributions and remarks from a Portuguese amateur of human genetics, regarding the genetic distance of different Han Chinese populations and other East Asian / SE Asian populations.

His original post is on Quora, here's the link:

https://www.quora.com/Are-the-Han-Chinese-very-similar-to-each-other-genetically-and-who-are-they-most-closely-related-and-or-similar-regarding-their-genetic-ancestry

He concluded that on a world scale Han Chinese and East Asians seem to be rather homogeneous, but on a regional scale the diversity actually isn't that bad. He said that according to his analysis the largest genetic distance observed among Han Chinese is the one between Han_Shanxi and Han_Guangdong at 0.08754, and the distance is comparable to the one between Portuguese and Danish at 0.08705. He further added that Han_Guangdong is the most divergent among all Hans and is closer to Vietnamese, and that all Hans exist on a cline from Koreans to Vietnamese.

It's unfortunate that he doesn't have access to samples of Guangxi Han, cause according to some previous studies and analyses Guangxi Han are the most divergent out of all Hans, even more divergent than those from Guangdong. He also doesn't have access to samples from Guizhou and Yunnan, cause some of them can be rather divergent as well.

Anyways, here're the tables and graphs that he presented.

https://qph.fs.quoracdn.net/main-qimg-cbaf2d4043f3ba04346ab9ec5575e311

https://qph.fs.quoracdn.net/main-qimg-1237e042642acc38f74707ed0aac0d21

https://qph.fs.quoracdn.net/main-qimg-10e77f83e269ab834f7bf9cefb53a972

https://qph.fs.quoracdn.net/main-qimg-be095b42b73ca596e4ad60f003b03c9c

https://qph.fs.quoracdn.net/main-qimg-8e6752ae547df9daa2e63d331876bda2

https://qph.fs.quoracdn.net/main-qimg-966e7ad80c5ac29bfccdfb6e6beccd83

https://qph.fs.quoracdn.net/main-qimg-a23de62e3a73bd62c6df5ca2d0b64e22

okarinaofsteiner
11-27-2021, 06:06 PM
Very interesting contributions and remarks from a Portuguese amateur of human genetics, regarding the genetic distance of different Han Chinese populations and other East Asian / SE Asian populations.

His original post is on Quora, here's the link:

https://www.quora.com/Are-the-Han-Chinese-very-similar-to-each-other-genetically-and-who-are-they-most-closely-related-and-or-similar-regarding-their-genetic-ancestry

He concluded that on a world scale Han Chinese and East Asians seem to be rather homogeneous, but on a regional scale the diversity actually isn't that bad. He said that according to his analysis the largest genetic distance observed among Han Chinese is the one between Han_Shanxi and Han_Guangdong at 0.08754, and the distance is comparable to the one between Portuguese and Danish at 0.08705. He further added that Han_Guangdong is the most divergent among all Hans and is closer to Vietnamese, and that all Hans exist on a cline from Koreans to Vietnamese.

It's unfortunate that he doesn't have access to samples of Guangxi Han, cause according to some previous studies and analyses Guangxi Han are the most divergent out of all Hans, even more divergent than those from Guangdong. He also doesn't have access to samples from Guizhou and Yunnan, cause some of them can be rather divergent as well.

-snip-

I saw this post too- this person is Brazilian by the way. I agree that the G25 province samples are rather limited. I wonder if this person saw my MDLP K23b graphs on how Koreans, Han Chinese, and Vietnamese score on MDLP K23b Austronesian and Tungus_Altaic which also show the same thing.

I’m willing to bet that Qinghai and Gansu Han are fairly divergent from other Northern Han- possibly to the same extent as Guangxi from Guangdong, or from Guangdong to Fujian- if not more.

Max_H
11-27-2021, 08:03 PM
I saw this post too- this person is Brazilian by the way. I agree that the G25 province samples are rather limited. I wonder if this person saw my MDLP K23b graphs on how Koreans, Han Chinese, and Vietnamese score on MDLP K23b Austronesian and Tungus_Altaic which also show the same thing.

I’m willing to bet that Qinghai and Gansu Han are fairly divergent from other Northern Han- possibly to the same extent as Guangxi from Guangdong, or from Guangdong to Fujian- if not more.

I am curious as to what pulls Shandong Han that much toward Tibetans... You would think they would be part of the Guangdong-Korea cline instead. I have seen Shandong Han score various more "western" components such as Tibetan-related or even low West Eurasian (Shanxi and xibei Han overall score a lot more of that even if still low overall) but they appear a lot more "eastern/Korean-shifted" than Han from Shanxi or even Henan.

I agree with you many Qinghai and Gansu Han are fairly divergent, but these areas have also recently experienced high migration from further east and even south from what I know... However, unlike Guangxi Han Gansu and Qinghai Han will mostly be divergent due to increased Tibetan and West Eurasian-like components while in Guangxi ofc you have mostly increased southern components.

okarinaofsteiner
11-28-2021, 12:17 AM
I am curious as to what pulls Shandong Han that much toward Tibetans... You would think they would be part of the Guangdong-Korea cline instead. I have seen Shandong Han score various more "western" components such as Tibetan-related or even low West Eurasian (Shanxi and xibei Han overall score a lot more of that even if still low overall) but they appear a lot more "eastern/Korean-shifted" than Han from Shanxi or even Henan.

I agree with you many Qinghai and Gansu Han are fairly divergent, but these areas have also recently experienced high migration from further east and even south from what I know... However, unlike Guangxi Han Gansu and Qinghai Han will mostly be divergent due to increased Tibetan and West Eurasian-like components while in Guangxi ofc you have mostly increased southern components.

Actual Northern Han from north of the Qinling-Huaihe line are genetically quite homogeneous, so if one (non-Qinghai/Gansu) northern province is Tibetan-shifted or Central Asian-shifted then they all are. Also genotype =/= phenotype, and the G25 samples may suffer from small sample size issues.

The "Guangdong Han to Korean/Japanese" cline shows that Fujian Han are noticeably more southern shifted compared to Hubei Han, and also Chongqing Han, who actually seem to be on a Vietnamese to Korean/Japanese cline. "Chongqing Han" is more "Japanese + Vietnamese" while "Sichuan Han" is more "Korean + Vietnamese".

MNOPSC1b
11-28-2021, 04:05 PM
The "Guangdong Han to Korean/Japanese" cline shows that Fujian Han are noticeably more southern shifted compared to Hubei Han, and also Chongqing Han, who actually seem to be on a Vietnamese to Korean/Japanese cline. "Chongqing Han" is more "Japanese + Vietnamese" while "Sichuan Han" is more "Korean + Vietnamese".

Fujian Han aren't noticeably more southern-shifted than Chongqing Han, the two are at a similar level in terms of the north-south parameter, their only difference lies in the east-west parameter.

Although you love to group Fujian Han with Guangdong "Han", from the above analyses it's clear that Guangdong "Han" are noticeably more southern-shifted than Fujian Han. And I expect Guangxi "Han" to be even more southern-shifted and closer to SE Asians. Guangdong and Guangxi "Han" aren't really Han from a genetic POV, but rather they are sinicized Tai-Kradai.

Ajeje Brazorf
12-03-2021, 06:24 PM
What are the ancestral components of Eastern Eurasians? In Western ones we have EEF, WHG, CHG, and ANE: are there similar components for Asians?

okarinaofsteiner
12-03-2021, 10:28 PM
What are the ancestral components of Eastern Eurasians? In Western ones we have EEF, WHG, CHG, and ANE: are there similar components for Asians?

very comprehensive summary (https://genetichistoryofeastasians.quora.com/The-recent-findings-of-academic-studies-from-2021-including-archeologic-linguistic-and-genetic-research#comments) (not by me)


Among noteworthy branches:


Andamanese (samplified by the Onge) and Hoabinhians split from the common ancestor of East Asians between 40,000 to 27,000 years ago.
Ancestral Native Americans, split between 36,000 to 23,000 years ago.
Jōmon people of Japan and Himalayans (Ancient Tibetans) split between 25,000 to 15,000 years ago.
Paleo-Siberians split between 22,000 to 18,000 years ago.
Northern East Asian and Southern East Asian split with each other about 12,000 years ago.


For East Asians proper, probably Paleo-Siberian, Amur Neolithic, Yellow River Neolithic (millet-farmer), and Yangtze Neolithic (rice-farmer). The last 3 seem to cluster with each other, although Yellow River and Yangtze River split from each other sometime during the last Ice Age.

There are also other ancestries that only certain East Asian populations have, which the above link calls “Basal East Asian”. We have Jomon, Hoabinhian/Onge, and Tianyuan- the last of which is an extinct lineage. There is also a diverged ancestry component found in Tibetans, although idk what this group’s relationship is with “Basal East Eurasian”, “Sahul”, and AASI.

Philippine Negritos seem to be a mix between “Basal East Eurasian” and Sahul “South Eurasian”.

Modern-day Papuans are mostly “Sahul” but seem mixed with “Basal East Eurasian” and “Austronesian” rice farmer ancestry (“East Asian proper”).

Some Siberian and Central Asian groups may have some sort of ANE-like ancestry or admixture? I know modern-day Mongols have some West Eurasian ancestry. Many modern-day SE Asian groups have some South Asian (steppe + “Iran farmer” + actual AASI) ancestry on top of Hoabinhian.

Ajeje Brazorf
12-03-2021, 10:54 PM
very comprehensive summary (https://genetichistoryofeastasians.quora.com/The-recent-findings-of-academic-studies-from-2021-including-archeologic-linguistic-and-genetic-research#comments) (not by me)



For East Asians proper, probably Paleo-Siberian, Amur Neolithic, Yellow River Neolithic (millet-farmer), and Yangtze Neolithic (rice-farmer). The last 3 seem to cluster with each other, although Yellow River and Yangtze River split from each other sometime during the last Ice Age.

There are also other ancestries that only certain East Asian populations have, which the above link calls “Basal East Asian”. We have Jomon, Hoabinhian/Onge, and Tianyuan- the last of which is an extinct lineage. There is also a diverged ancestry component found in Tibetans, although idk what this group’s relationship is with “Basal East Eurasian”, “Sahul”, and AASI.

Philippine Negritos seem to be a mix between “Basal East Eurasian” and Sahul “South Eurasian”.

Modern-day Papuans are mostly “Sahul” but seem mixed with “Basal East Eurasian” and “Austronesian” rice farmer ancestry (“East Asian proper”).

Some Siberian and Central Asian groups may have some sort of ANE-like ancestry or admixture? I know modern-day Mongols have some West Eurasian ancestry. Many modern-day SE Asian groups have some South Asian (steppe + “Iran farmer” + actual AASI) ancestry on top of Hoabinhian.

Thank you for your response. These are some of the samples I normally use to make models, but the picture is not complete because there is a lot of ancient DNA missing from Asia.


Ami:NA13615,0.023903,-0.444802,-0.048649,-0.074613,0.14772,0.069723,0,-0.005077,-0.025565,-0.020046,0.049041,0.006444,-0.01219,-0.007844,0.006243,0.006762,0.006258,0.005194,0.007 29,-0.027263,0.008485,-0.029306,0.001356,-0.004338,-0.04778
CHN_Western_Liao_River_BA_o:91KLM2,0.022765,-0.468159,0.093149,-0.044251,-0.088632,-0.05271,0.013866,0.025614,0.006749,0.020046,-0.039948,0.001499,-0.006244,0.021882,0.003122,-0.002917,0.001434,-0.003674,0.01081,0.034016,-0.029822,-0.029306,-0.050408,-0.005784,-0.003832
Even:even2682,0.042115,-0.442771,0.144814,-0.019057,-0.142488,-0.086735,0.028201,0.038306,0.035587,0.024966,0.062 357,0.010041,-0.011596,0.003303,-0.026058,-0.017104,-0.000522,0.017483,0.040726,0.018009,0.058896,-0.062445,0.005916,0.014098,0.026824
JPN_Nagabaka_historic:NAG038,0.014797,-0.343249,-0.050157,0.013566,0.038469,0.008925,-0.004935,-0.002077,0.022089,0.02442,-0.051315,-0.004496,0.00996,-0.016377,-0.021851,-0.013524,0.006258,0.014189,0.016215,-0.012506,0.064511,-0.052305,0.009737,0.006025,-0.112206
Ket:584_R01C01,0.079676,-0.247789,0.126335,0.052326,-0.107097,-0.029562,0.00188,0.012461,0.002863,-0.028976,0.067878,0.00045,0.020515,-0.10886,-0.024837,-0.014187,0.005998,0.003927,0.014078,-0.002626,-0.028949,0.052676,0.069265,-0.029643,-0.020238
Mari:mari5,0.097888,-0.04773,0.092395,0.060078,-0.028005,0.001394,0.012456,0.019153,-0.006136,-0.041185,0.026307,-0.019483,0.041476,-0.044177,-0.037866,-0.016971,0.000913,-0.008361,-0.037081,-0.029889,0.01959,0.003957,-0.055831,0.010363,-0.003712
MYS_LN:Ma912,0,-0.395041,-0.084098,-0.034561,0.154183,0.068049,-0.007285,-0.010615,-0.005522,-0.018224,0.090938,0.017834,-0.015163,0.006331,0.006922,-0.001061,-0.001695,-0.006841,-0.003394,0.019259,-0.01697,0.012984,-0.009737,0.00723,0.048379
Paniya:PY-6,0.012521,-0.169593,-0.188183,0.142444,-0.049548,0.057173,-0.008225,0.017538,0.096944,0.059409,0.008931,0.001 049,-0.002527,0.018579,-0.025651,-0.03739,0.005737,0.002154,0.001006,0.031015,0.0061 14,0.01422,-0.012078,0.00482,-0.008263
Qiang_Danba:DBA01,0.020488,-0.438709,0.006034,-0.05168,0.023081,0.006693,0.004935,0.001154,-0.009613,0.012028,-0.097595,-0.005095,0.009068,-0.010184,-0.012351,-0.001591,0.000782,-0.00152,-0.015838,0.004502,0.005116,0.033015,0.017008,-0.006868,0.037721
RUS_AfontovaGora3:AfontovaGora3,0.093335,-0.01828,0.083344,0.231592,-0.091402,0.042949,-0.063688,-0.078228,-0.035383,-0.096221,0.047417,-0.010491,0.023637,-0.074454,0.020358,0.02254,-0.012908,0.003801,-0.003394,0.000625,-0.03706,0.015209,0.013927,0.009278,-0.005149
Surui:HGDP00843,0.053497,-0.311768,0.11653,0.11079,-0.119407,-0.026216,-0.33489,-0.398983,-0.019839,-0.022962,0.006496,-0.007943,0.005798,0.039635,-0.016151,0.000133,0.009127,-0.00266,-0.006788,0.001,-0.004118,0.012365,-0.010599,-0.00976,-0.007185
Yakut:455_A,0.044391,-0.378792,0.112759,-0.023579,-0.107097,-0.071954,0.029141,0.034845,0.026384,0.024602,0.034 914,0.015586,-0.025867,0.074454,0.045738,0.024927,0.001565,-0.01761,-0.042863,-0.033641,0.000624,0.039198,0.011092,-0.001325,-0.019759

MNOPSC1b
12-04-2021, 01:55 AM
What are the ancestral components of Eastern Eurasians? In Western ones we have EEF, WHG, CHG, and ANE: are there similar components for Asians?

I know okarina had already answered and provided a link, but I'd also like to share what I know, according to the papers I've read.

For the sake of simplicity I won't talk about peoples like Ust-Ishim, Oase, and Bacho Kiro, although it's very likely they were related to East Eurasians, their respective positions within the East Eurasian tree are still not well understood and remain somewhat controversial.

The first group that split off from East Eurasians were likely the ancestors of Papuans and Australian aborigines, they likely diverged some 50,000 to 45,000 years BP, and intermixed with Denisovans (direct ancestors of East Asians also intermixed with Denisovans, but not to the same extent as Papuans and Australian aborigines).

Hoabinhians / Onge likely split off some 40,000 years ago, along with ancestors of Longlin, Tianyuan, and Amur33K. They likely represent a very basal East Eurasian lineage.

Shortly after the Jomon split off, likely sometime between 40,000 to 30,000 years BP.

Then the ancestors of Native Americans split off around 27,000 years BP, and got intermixed with ANE. Native Americans can be modeled as roughly 70% East Eurasian + 30% ANE.

And finally the ancestors of Northern East Asian and Southern East Asian split off around 23,000 years BP.

For East Asian proper like okarina said, three components are dominant, Amur Neolithic, Yellow River Neolithic, and Yangtse Neolithic. Amur Neolithic peaks among Tungusic tribes, Mongols, and Koreans/Japanese; Yellow River Neolithic is the predominant component among modern East Asians and can be found all around from north to south, but peaks among certain Tibeto-Burman tribes, Northern Chinese, and as well as Koreans/Japanese; Yangtse Neolithic peaks among Taiwanese aborigines, but also reaches significant proportions among Far South Chinese, South Chinese minorities, and SE Asians.

Other minor components include Paleo-Siberian, which is mostly found among Siberian peoples like Yakut, Chukchi, etc. There's another minor component that I would label as Austroasiatic hunter-gatherers, which is probably an admixed component of Hoabinhians with Neolithic Yangtse. This component peaks among SE Asian tribal peoples like Mlabri and Htin, but can also be found among most SE Asians, and certain South Chinese minorities. Many SE Asians also have some West Eurasian ancestry that they likely acquired through contacts with South Asians.

This is my understanding of what has been shown to us in the papers so far published, feel free to correct any errors I've made.

And also I suggest be cautious with the Quora link that okarina provided, don't read too much into what the Saito Takashi guy wrote. I've debated with that guy on Quora before and I don't have a very good impression about him. He's a very stubborn supporter of the southern route hypothesis, even though we don't have that much evidence for that as of now. I'm not saying that the papers and sources he quotes are necessarily wrong, but his subjective interpretation of his sources can be quite misleading at times.

okarinaofsteiner
12-28-2021, 09:46 PM
A former Anthroscape member sent me a chart of 23mofang averages for various suburban and rural districts in China, which are believed to be more autosomally "representative" of specific regions and linguistic subgroups (as opposed to the city centers, which are more cosmopolitan and therefore more "mixed".) https://imgur.com/a/pxuhh3D

https://i.imgur.com/1UYN0Tn.jpg

Map (latitude + longitude) with ChinaMAP study clusters added
https://i.imgur.com/GX11i2q.png


I just saw some GEDmatch results for a Hui person (both parents are Hui from China proper (https://np.reddit.com/r/23andme/comments/py5tjs/hui_chinese_results_photo_maternal_haplo_c4a/hes05mm/)) living outside China, thought I'd share them in this thread too for further analysis:

MDLP K23b (https://np.reddit.com/r/23andme/comments/q1fvw5/gedmatch_results/): 44.65% S_EA, 32.91% T_A, 17.24% AN, 0.95% E_Sib [95.75% East Asian] + 0.46% Amerind + 0.86% S_CentralAsian, 1.66% EEF, 1.26% N_Afr [3.78% West Eurasian].

HarappaWorld: 75.18% NE Asian, 11.25% SE Asian, 5.83% Siberian, 2.41% Baloch, 1.69% American, 1.52% SW Asian, 0.98% Beringian, 0.5% NE Euro, 0.3% Pygmy, 0.21% Mediterranean, 0.12% W African

Dodecad K12b: 65.88% East Asian, 22.51% SE Asian, 5.86% Siberian, 2.27% N Euro, 1.61% Gedrosia, 1.48% NW African



Re: MDLP K23b- Those Tungus_Altaic and Austronesian numbers are very close to my Simulated Jin speaker averages (dark red dots), although their East Eurasian percentage is 1-2% lower than the simulated Jin averages.

(original AN vs T_A plot)
https://i.imgur.com/SE1mMcf.png

(original N-S cline vs % East Asian plot)
https://i.imgur.com/FhIURm2.png

This person scores 0.40825 on my N-S cline. Given that they score a little under 96% East Asian, this puts them in the range of the most northern-shifted and most-West Eurasian shifted Chinese samples in my original dataset.

(graph does not include this data point)
https://i.imgur.com/FhIURm2.png

For context, their 23andMe results were

96.0% East Asian/Amerindian (69.9% Chinese from Shandong, Jiangsu, Guangdong, Henan, and apparently Taiwan; 14.8% Mongolian/Manchurian, 8.9% Korean, 2.4% Broadly East Asian),
1.6% West Asian/North African (0.8% Iran + Caucasus + Mesopotamia, 0.8% General),
1.1% Central + South Asian (0.5% Bengali/NE Indian, 0.2% Central Asian, 0.3% South Indian/Sri Lankan, 0.1% General),
0.2% Trace Spanish/Portuguese,
1.1% Unassigned.

okarinaofsteiner
01-15-2022, 09:51 PM
Taiwanese American (https://np.reddit.com/r/23andme/comments/o60mva/taiwanese_american_anyone_know_if_there_is/h2qiepb/), half waishengren half benshengren (https://np.reddit.com/r/23andme/comments/o60mva/taiwanese_american_anyone_know_if_there_is/h2rgdg1/). Scored 100% "Chinese" on 23andMe, with Guangdong and Taipei as the top region matches.

This person is 47.2% S_EA, 30.33% AN, 21.38% T_A, 98.91% East Asian, and scores 0.545 on my North-South cline. A bit more northern-shifted than most Fujian Han but not completely outside the Fujian range it seems. This person also has some Hoabinhian-like trace (<1% unless we interpret the "Amerindian" noise as such) ancestry, which isn't unusual for South China Han. Although it would be interesting to find out whether it's from the waishengren and/or benshengren parents.

MDLP K23b Oracle results:


Admix Results (sorted):
# Population Percent
1 South_East_Asian 47.2
2 Austronesian 30.33
3 Tungus-Altaic 21.38
4 Amerindian 0.61
5 South_Indian 0.32
6 Australoid 0.15

Single Population Sharing:
# Population (source) Distance
1 Hakka ( ) 3.36
2 Chinese_Taiwan ( ) 4.4
3 Han_Singapore ( ) 4.62
4 Han ( ) 5.8
5 Tujia ( ) 7.16
6 Han-Mandarin ( ) 7.9
7 Jinuo ( ) 7.96
8 She ( ) 9.79
9 Hmong_Miao ( ) 9.98
10 Cantonese ( ) 10.27
11 Lawa ( ) 10.68
12 Paluang ( ) 11.39
13 Yao ( ) 11.51
14 Miao ( ) 11.69
15 Hmong ( ) 12.27
16 Wa ( ) 13.18
17 Karen ( ) 13.61
18 Plang ( ) 15.25
19 Han_North ( ) 15.88
20 Tai_Yuan ( ) 16.21

Mixed Mode Population Sharing:
# Primary Population (source) Secondary Population (source) Distance
1 84.1% Han ( ) + 15.9% Japanese ( ) @ 0.77
2 81.9% Han ( ) + 18.1% Ryukyuan ( ) @ 0.83
3 80.9% Han ( ) + 19.1% Japanese_ML ( ) @ 0.88
4 75.7% She ( ) + 24.3% Japanese ( ) @ 1.24
5 72.7% She ( ) + 27.3% Ryukyuan ( ) @ 1.26
6 77% Han ( ) + 23% Korean_KR ( ) @ 1.29
7 69.3% Hmong_Miao ( ) + 30.7% Korean ( ) @ 1.33
8 89% Chinese_Taiwan ( ) + 11% Naga ( ) @ 1.34
9 67.5% Naga ( ) + 32.5% Atayal_Coriell ( ) @ 1.36
10 83.8% Hakka ( ) + 16.2% Han_North ( ) @ 1.4
11 91.7% Hakka ( ) + 8.3% Naga ( ) @ 1.43
12 71.5% She ( ) + 28.5% Japanese_ML ( ) @ 1.48
13 89.1% Chinese_Taiwan ( ) + 10.9% Nysha ( ) @ 1.51
14 91.1% Hakka ( ) + 8.9% Tibetian_TTR ( ) @ 1.52
15 85.3% Chinese_Taiwan ( ) + 14.7% Naxi ( ) @ 1.52
16 89.7% Chinese_Taiwan ( ) + 10.3% Aonaga ( ) @ 1.55
17 91.2% Hakka ( ) + 8.8% Tibetian_Madou ( ) @ 1.56
18 91.9% Hakka ( ) + 8.1% Nysha ( ) @ 1.56
19 69.1% Naga ( ) + 30.9% Igorot ( ) @ 1.58
20 85.3% Chinese_Taiwan ( ) + 14.7% Yi ( ) @ 1.58

Shuzam87
01-15-2022, 10:35 PM
My MDLP K23b:
Admix Results (sorted):
#Population Percent
1. South_East_Asian 44.97
2. Tungus-Altaic 27.77
3. Austronesian 25.32
4. European_Early_Farmers 0.79
5. Caucasian 0.35
6. Melano_Polynesian 0.32
7. European_Hunters_Gatherers 0.24
8. North_African 0.15
9. Arctic 0.09

My mom's MDLP K23b:
Admix Results (sorted):
# Population Percent
1. South_East_Asian 43.71
2. Tungus-Altaic 29.62
3. Austronesian 20.05
4. Caucasian 2.15
5. East_Siberian 1.82
6. Near_East 1.35
7. Arctic 0.63
8. European_Hunters_Gatherers 0.44
9. Khoisan 0.18
10. Archaic_Human 0.05
11. North_African 0.01

My grandpa's MDLP K23b:
Admix Results (sorted):
# Population Percent
1. South_East_Asian 45.23
2. Tungus-Altaic 31.5
3. Austronesian 19.13
4. East_Siberian 2.42
5. Arctic 1.45
6. East_African 0.28

My K12b:
Admix Results (sorted):
#Population Percent
1. East_Asian 60.37
2. Southeast_Asian 32.12
3. Siberian 4.06
4. Caucasus. 1.91
5. Atlantic_Med 0.77
6. North_European 0.5
7. Sub_Saharan 0.15
8. East_African 0.12

My mom‘s K12b:
Admix Results (sorted):
# Population Percent
1. East_Asian 60.8
2. Southeast_Asian 28.32
3. Siberian 5.84
4. Gedrosia 1.49
5. Southwest_Asian 1.33
6. Caucasus 0.99
7. East_African 0.64
8. North_European 0.47
9. Sub_Saharan 0.12

My grandpa's K12b:
Admix Results (sorted):
# Population Percent
1. East_Asian 65.43
2. Southeast_Asian 26.87
3. Siberian 7.34
4. Gedrosia 0.25
5. East_African 0.11

Ajeje Brazorf
01-17-2022, 10:39 PM
East Asian-related PCA

Samples included
Part 1: https://pastebin.com/raw/H0AA9XtE
Part 2: https://pastebin.com/raw/bxjci2MV

https://i.imgur.com/xlmnKW3.png
https://i.imgur.com/DoAPw1x.png
https://i.imgur.com/4EfikSj.png
https://i.imgur.com/3V93vqY.png
https://i.imgur.com/bkPFjJN.png