MitchellSince1893
09-23-2014, 05:20 AM
It occurred to me that 23andme customers could use their ~1000 matches as somewhat random polling data to get an idea of the percentage of y-dna (or m-dna) in this rather US centric database. In other words you could get an idea of the make up of y-dna or mitochondrial dna in the US
For example assuming half the matches are male:
-Of my approx. 553 male matches 194 (35%) are R1b-P312, 65 are R1b-U152 (12%) and 40 (7.2%) are R1b-L2
-Of my mother's approx 533 male matches 165 (31%) are P312, 37 are U152 (7%), and 20 (3.7%) are L2
-Of my father's approx 505 male matches 188 (37%) are P312, 52 (10%) are U152, and 31 (6.1%) are L2.
In this small sample of kits it appears P312 makes up ~35% of 23andme's database, while U152 is ~28% of P312, and L2 is ~58% of U152.
More samples would give you a higher confidence and if you had enough participants you could get a pretty good idea of the actual 23andme database percentages.
An example of how this could be used in the forums would be a member starting a new thread in say the L21 section asking 23andme customers to post their percentage of L21 matches.
For example assuming half the matches are male:
-Of my approx. 553 male matches 194 (35%) are R1b-P312, 65 are R1b-U152 (12%) and 40 (7.2%) are R1b-L2
-Of my mother's approx 533 male matches 165 (31%) are P312, 37 are U152 (7%), and 20 (3.7%) are L2
-Of my father's approx 505 male matches 188 (37%) are P312, 52 (10%) are U152, and 31 (6.1%) are L2.
In this small sample of kits it appears P312 makes up ~35% of 23andme's database, while U152 is ~28% of P312, and L2 is ~58% of U152.
More samples would give you a higher confidence and if you had enough participants you could get a pretty good idea of the actual 23andme database percentages.
An example of how this could be used in the forums would be a member starting a new thread in say the L21 section asking 23andme customers to post their percentage of L21 matches.