Post your Marques K64

[Arlus]

Target: Arlus_scaled
Distance: 1.6738% / 0.01673789 | ADC: 0.25x
94.8 South_central_Asia
4.6 South_Asian
0.6 Amazonian

Target: Arlus_scaled
Distance: 1.6607% / 0.01660733
98.4 South_central_Asia
0.8 Amazonian
0.6 Indonesian_Austronesian
0.2 Greece

[Mis]

Target: Marek_scaled
Distance: 1.5714% / 0.01571437 | ADC: 0.25x
80.6 Central_European
11.6 Central_Asia_and_steppe
2.8 East_European
2.4 Baltic
1.6 USA_European_Settler
0.8 East_Africa
0.2 Levantine

Target: Marek_scaled
Distance: 1.1217% / 0.01121687
45.8 Baltic
20.2 West_Russia
13.4 Balkan
12.8 East_European
3.6 Levantine
2.8 USA_European_Settler
0.6 Anatolia
0.6 Central_European
0.2 East_Africa

[agent_lime]

What's fascinating is the lack of Persian_Babylonian in some of the runs from persons further East into the Indian subcontinent than the confines of Pakistan and Western India.

They don't have direct ingress from Persians. Only NW South Asians + some Western Pakistanis show Iran_chl mixture. Especially with the "base" populations covering the Iran_N.

[DMXX]

IMHO we're reaching a saturation point with all of these custom G25-based calculators as it is.

Particularly as several of them fall prey to the (unintentional or otherwise) biases or knowledge deficits of their creators.

I don't see a rationale in merging "Persia" (I assume this is the entire Iranian plateau) and "Mesopotamia" into a single entity given the inherent differences between them (let alone within Iran itself), whilst giving "Mediterranean islanders" their own exclusive component.

Makes as much sense as merging NW S Asians with interior Subcontinental pops (f.ex. a "Kalash-Hyderabad" cluster). Sure, there's clear overlap in the agriculturalist component between W Iran and Mesopotamia, but Assyrians and Mazandarani or Bandari Iranians are pretty darn divergent.

We'd be much better off as a community if people began to independently run and appraise qpAdm models, IBD data, downstream uniparentals and the like rather than a seemingly endless wave of custom arrangements from multiple chefs carving out the same cake in different ways.

[agent_lime]

IMHO we're reaching a saturation point with all of these custom G25-based calculators as it is.

Particularly as several of them fall prey to the (unintentional or otherwise) biases or knowledge deficits of their creators.

I don't see a rationale in merging "Persia" (I assume this is the entire Iranian plateau) and "Mesopotamia" into a single entity given the inherent differences between them (let alone within Iran itself), whilst giving "Mediterranean islanders" their own exclusive component.

Makes as much sense as merging NW S Asians with interior Subcontinental pops (f.ex. a "Kalash-Hyderabad" cluster). Sure, there's clear overlap in the agriculturalist component between W Iran and Mesopotamia, but Assyrians and Mazandarani or Bandari Iranians are pretty darn divergent.

We'd be much better off as a community if people began to independently run and appraise qpAdm models, IBD data, downstream uniparentals and the like rather than a seemingly endless wave of custom arrangements from multiple chefs carving out the same cake in different ways.

I agree. However, I have some counterpoints.
- Allows new users to get some idea of their ancestry.
- Encourages more folks to get a G25.
- Is good to catch components that are not usual for ethnicity. Like for me I only get good fits if I can get some African, Native American, Arab and a Corded Ware heavy population. (I show roughly 2% MENA and 0.5 Somali on 23and me/ Used to get 2% Native American on Ancestry likely from ANE).
- Gets more folks interested in diving deep and running their own stuff.

I think a disclaimer to all is that these runs should not be taken seriously.

[DMXX]

Don't disagree - The over-arching response to those points (and my main gripe) is that the component selection procedure in most of these calculators is almost always at the whim of the author - Multiply that by two dozen and we have multiple calculators with varying degrees of appropriateness that'll either confuse newer users or feed into the preconceived biases of established ones.

As we both know, G25 data is basically one big exercise in Euclidean distance approximation in an estimated vs. observed set of coords. If custom calc creators arbitrarily define components, it'd either "accidentally" capture particular variants of ancestry through them, or dilute real signals. It'd just add more confusion instead of clarity.

For instance, this "Persia-Mesopotamia" component would ostensibly represent some sort of Iran_ChL-type (Anatolia_N-rich) admix in groups like NW S Asians. However, there was already substantial diversity in Iran_N:Anatolia_N proportions in Iran_ChL pops (let alone modern groups, who have substantial admix from other areas).
So, would a NW S Asian scoring X% of "P-M" in such a calc reflect an actual excess of Anatolia_N-rich Iran_ChL via admix from further west, or is it a false signal from another ancestral source (f.ex. a steppe-admixed Sappali Tepe, which is roughly what central modern Iranians are)?

There's no consensus formation regarding validation, cross-comparison for accuracy, or anything else that resembles a statistical assessment of how "useful" these calcs are versus a conventional regional G25 setup (like the ones we've performed robustly and as a community here).

I'm also not opposed to Davidski generating further income from more G25 (frankly, IMO, he deserves it for all the hard work he's done for nearly a decade).

[T90]

Scaled

Target:1


Distance: 1.0454% / 0.01045382
58.2	South_central_Asia
23.2	South_Asian
10.6	Central_Asia_and_steppe
2.6	Sibérian
2.0	Germanic_Europe
1.4	West_Russia
1.0	South_Caucasus
0.4	East_European
0.4	French
0.2	West_Berber


Target: 2


Distance: 1.8515% / 0.01851469
41.6	South_central_Asia
27.2	Central_Asia_and_steppe
13.4	South_Asian
6.4	Baltic
3.2	Italian
3.0	Indochina
2.8	Persian_Mesopotamian
1.4	USA_indigenous
0.6	Melanesian
0.4	Egyptian

[khanabadoshi]

Strange...

Target: Khana
Distance: 0.7369% / 0.00736941
36.0 Persian_Mesopotamian
28.6 South_Asian
17.0 South_central_Asia
5.4 Central_Asia_and_steppe
5.4 Greece
4.6 Sibérian
1.8 Crusaders
0.8 Baltic
0.4 Germanic_Europe

Target: Khana_Uncle
Distance: 1.1315% / 0.01131534
36.2 South_Asian
23.6 Central_Asia_and_steppe
20.0 South_central_Asia
9.8 Greece
6.8 Persian_Mesopotamian
1.4 Indo_Tibétan
1.0 Melanesian
0.8 Levantine
0.2 Baltic
0.2 Egyptian

Target: Khana_GF_Sister
Distance: 2.4832% / 0.02483166
54.2 South_central_Asia
18.4 Central_Asia_and_steppe
13.8 Persian_Mesopotamian
4.0 Egyptian
3.0 Spain
2.2 South_Asian
1.8 Italian
1.2 British
0.6 Sibérian
0.4 East_Africa
0.4 East_European

Target: Khana_GM
Distance: 0.8115% / 0.00811504
41.0 South_central_Asia
29.6 Central_Asia_and_steppe
5.6 Greece
5.0 East_European
5.0 Mongolian
4.8 Indo_Tibétan
4.0 Persian_Mesopotamian
3.2 Anatolia
1.6 Central_European
0.2 Balkan

Target: Khana_GM_Brother
Distance: 0.2753% / 0.00275284
34.8 South_central_Asia
25.0 Central_Asia_and_steppe
11.2 Anatolia
8.2 Indo_Tibétan
5.8 Persian_Mesopotamian
3.4 French
3.0 Balkan
2.2 Artic
1.4 Central_Chinese
1.4 Mongolian
1.0 South_Caucasus
0.8 Japanese/Korean
0.6 Central_European
0.6 Sibérian
0.4 Scandinavia
0.2 West_Russia

[ThaYamamoto]

Fits are much tighter with p much any calc with ancestrydna

[MonkeyDLuffy]

Target: Ramgarhia_Average
Distance: 1.1311% / 0.01131138
40.2 South_central_Asia
30.8 South_Asian
17.6 Persian_Mesopotamian
7.6 West_Russia
3.8 East_European

Ramgarhia_PV (new)

Target: Ramgarhia_PV_scaled
Distance: 1.0715% / 0.01071488
42.8 South_central_Asia
30.4 South_Asian
8.8 Persian_Mesopotamian
7.4 East_European
6.6 Anatolia
3.2 West_Russia
0.6 Central_Asia_and_steppe
0.2 Amazonian