We have covered several biohacking stories related to home DNA testing.
- Getting More from Home DNA Testing
- Nutrigenomics – Personalized Vitamin Supplements Based on DNA
- Biohacking: Merging DNA Results from 23andMe and Ancestry
What we have not discussed is a way to determine which home DNA test is offering more useful information. Let’s focus on the two most popular services which now cost ~$80. Ancestry.com and 23andMe are both testing for about the same amount of SNPs (Single-nucleotide polymorphisms) ~650,000. The number of SNPs being tested will change annually as the Illumina chip used for sequencing is regularly upgraded.
Who is offering more Ancestry.com or 23andMe? Ancestry.com is currently offering significantly more RAW data with their service then 23andMe. Here is the breakdown:
- Ancestry.com: SNPs in SNPedia: 47,000
- Ancestry.com: SNPs in ClinVar: 25,000
- Ancestry.com: “proprietary” SNPs: 0
- 23andMe: SNPs in SNPedia: 24,674
- 23andMe: SNPs in ClinVar: 2,000
- 23andMe: “proprietary” SNPs: 17,829
The popular DNA reporting site “promethease” relies heavily on data from SNPedia and Clinvar. We can see a great deal more results from clinical studies (12x) and explanations of SNPs (2x) using a DNA report from Ancestry.com. 23andMe has been increasing their proprietary SNPs making it difficult for third parties to interpret all of the data they are collecting. It is worth mentioning that when using third parties for DNA report summaries that 23andMe is currently preferred and commonly the only format accepted.
One remaining question is how many SNPs do 23andMe and Ancestry.com have in common. Not being sure myself I wrote a simple python script to compare them. These files are quite large and this could actual use a serious algorithm doing better pattern matching and buffering to quickly match SNPs between files. My DNA SNP Compare script estimates that there are ~140,000 SNPs in common between the two services.
For comparing two lists of objects, and looking for objects that are in both, you want to use sets (as in set theory):
s1 = set(file1_snps_sort)
s2 = set(file2_snps_sort)
sboth = s1.intersection(s2)
print(len(sboth), “SNPs are in both sets”)