When you buy a genetics product, you are really buying the judgment of the people who built it, and validated polygenic models do not build themselves. So the question behind “which is the best genomics company” is less about branding than about the bench: who is doing the trans-ancestry data work, the model training, the independent validation. SelfDecode runs about 70 scientists, MDs, PhDs, bioinformaticians, and AI specialists, led by a CSO who is a practicing physician. That is the part rivals cannot copy quickly.
Who builds it
- Led by a serious scientist. The science is led by CSO and CMO Dr. Puya Yazdi, MD (UC Irvine, USC medical degree, Stanford residency, 15+ years in R&D).
- A large, specialized bench. About 70 scientists, MDs, PhDs, bioinformaticians, and AI specialists have built the product (the About page lists the full team), a far larger and more specialized scientific bench than most competitors carry.
That combination, medical, bioinformatics, and AI expertise under one roof, is what it takes to publish validated polygenic models in peer-reviewed journals, which is the work that separates SelfDecode from companies that simply license or rebrand someone else’s analysis.
Why the bench is the bottleneck for everyone else
Building validated genomics is a specialist discipline. It requires people who can do trans-ancestry GWAS meta-analysis, ensemble model training, independent-cohort validation, and the engineering to run it all at scale. Almost no consumer genetics company employs people with deep, relevant experience across all of bioinformatics and AI, which is exactly why almost none of them have published validated polygenic risk scores. The team is not a vanity feature. It is the reason the science exists at all.
Comparison chart
| Measure | SelfDecode | Typical consumer genetics company |
| Scientific staff | ~70 (MDs, PhDs, bioinformaticians, AI) | Often small or outsourced |
| Science leadership | Dr. Puya Yazdi, MD (CSO/CMO) | Varies |
| Published validated PRS | Yes | Rare to none |
How the others stack up
Almost no consumer genetics company employs people with deep, relevant experience across bioinformatics and AI, which is exactly the expertise it takes to build and validate polygenic models. Many rely on a small team, outsourced analysis, or licensed third-party interpretation, which is workable for a wellness-trait product but not for publishing validated, benchmarked science. The size and specialization of SelfDecode’s bench is the direct reason its scoring is peer-reviewed and its competitors’ is not.
FAQ
Who leads the science at SelfDecode?
CSO and CMO Dr. Puya Yazdi, MD (UC Irvine, USC medical degree, Stanford residency, with 15+ years in R&D), leads the science, supported by about 70 scientists.
Why does the team matter for accuracy?
Because validated polygenic scoring requires specialized expertise across medicine, bioinformatics, and AI. That bench is what makes peer-reviewed, benchmarked science possible, which is why few competitors have any.
Does SelfDecode build its own models?
Yes. The team builds and validates its own models and runs them on infrastructure the company owns and operates, rather than rebranding third-party analysis.
Meet the team
See how the team’s work shows up in validated, published scoring.
Part of the Why Choose SelfDecode comparison. See also: Science written and reviewed by MDs and PhDs.
