How many SNPs needed to be reliable on Gedmatch?

[Ljubic]

I just noticed that my fathers 23andme raw file only uses 58000 SNPs in the admixture calculator while my Ancestry file uses 178000!! Thats on Eurogenes K13 but some other even use less. At first i tought there might have been some faults while the raw file has been uploaded on Gedmatch so i deleted his old file and uploaded it again twice. The results nevertheless have been the same.

[Kaspias]

120k+ reliable imo

[Ljubic]

58000 is a low hanging fruit. The testing company fucked something up it seems. Maybe that explains his unusual results.

Gedmatch Kit: CR8737733

Edited the Gedmatch kit number in for reference.

[digital_noise]

58000 is a low hanging fruit. The testing company fucked something up it seems. Maybe that explains his unusual results.

Gedmatch Kit: CR8737733

Edited the Gedmatch kit number in for reference.

No, thats just 23 and Me's V5 chip. Not sure, but I think its due to imputation on the new chip so they actually test less SNP's. I could be wrong on the details but thats basically why there are only 58000 used.

My 23 and Me v5 chip gives different results in Gedmatch when compared to My Heritage, Ancestry and FTDNA which all give practically identical results.

[happycow]

V5 chip. convert the raw data to V3 and you will get more snp's.

[talombo]

This is what I get in Eurogenes K13:

FTDNA OmniExpress chip: 175029
LivingDNA Sirius chip: 130787
23andme v5: chip 58863

But the results don't differ much between them.

87488

[Ljubic]

The difference between 23andme and FTDNA is not gigantic but still noticeable. Could you post your oracles please?

[talombo]

The difference between 23andme and FTDNA is not gigantic but still noticeable. Could you post your oracles please?


FTDNA
Single Population Sharing:

# Population (source) Distance
1 Sephardic_Jewish 6.64
2 Algerian_Jewish 6.71
3 Tunisian_Jewish 7.17
4 Cyprian 7.57
5 Libyan_Jewish 7.82
6 Italian_Jewish 8.09
7 South_Italian 9.24
8 Lebanese_Muslim 11.24
9 East_Sicilian 11.33
10 Ashkenazi 12.03
11 Syrian 12.2
12 Central_Greek 12.82
13 Samaritan 13.38
14 West_Sicilian 14.47
15 Lebanese_Christian 14.76
16 Jordanian 15.4
17 Italian_Abruzzo 15.52
18 Palestinian 15.65
19 Lebanese_Druze 16.25
20 Turkish 18.01


LivingDNA
Single Population Sharing:

# Population (source) Distance
1 Sephardic_Jewish 5.6
2 Algerian_Jewish 6.55
3 Tunisian_Jewish 7.41
4 Italian_Jewish 7.57
5 South_Italian 7.95
6 Libyan_Jewish 8.12
7 Cyprian 9.6
8 East_Sicilian 10.07
9 Ashkenazi 11.1
10 Central_Greek 11.69
11 Lebanese_Muslim 12.41
12 West_Sicilian 12.73
13 Syrian 13.2
14 Italian_Abruzzo 13.75
15 Samaritan 15.28
16 Jordanian 16.38
17 Lebanese_Christian 16.6
18 Palestinian 16.64
19 Lebanese_Druze 17.38
20 Turkish 17.93


23andme
Single Population Sharing:

# Population (source) Distance
1 Algerian_Jewish 6.58
2 Sephardic_Jewish 6.82
3 Tunisian_Jewish 7.69
4 Cyprian 8.11
5 Libyan_Jewish 8.22
6 South_Italian 8.25
7 Italian_Jewish 8.28
8 East_Sicilian 9.56
9 Ashkenazi 10.22
10 Central_Greek 11.08
11 Lebanese_Muslim 11.13
12 Syrian 11.89
13 West_Sicilian 13.29
14 Samaritan 14.1
15 Italian_Abruzzo 14.17
16 Jordanian 15.15
17 Lebanese_Christian 15.52
18 Palestinian 15.71
19 Lebanese_Druze 16.56
20 Turkish 16.91

[Ljubic]

Yeah it really is pretty much the same even though the number of SNPs used is very different.

[Lucas]

Yeah it really is pretty much the same even though the number of SNPs used is very different.

Old FTDNA had 160 000 snps. New probably about 70000 - 80000.

[Ljubic]

Old FTDNA had 160 000 snps. New probably about 70000 - 80000.

I wonder why less SNPs are used with the new chips though.

[Lucas]

I wonder why less SNPs are used with the new chips though.

It is said it is cheaper.