2025-01-18
I took inspiration from this X post and I used Claude Code to analyze the raw genetic data (.csv) I downloaded from tellmegen.
The results showed that I have a strong predisposition towards worrying too much (who doesn’t?), type 2 diabetes (already knew), I’m lactose intolerant (had a hunch) and I can metabolize coffee really well (already knew since I drink 6 coffees per day and sleep like a baby).
Disclaimer: This is not medical advice. I’m just a guy who ran his genetic data through an AI. Consult a healthcare professional for any health-related decisions.
You can follow this guide
No.
I paid around 100 Euros for this test just to find out that my ancestors are from Romania and I have a high diabetes risk.
Regarding your ancestry: if you live in Europe it’s really boring to see that your ancestors came from the region you were born into.
Regarding predispositions to diseases: sleep well, eat clean (low carb in my case) exercise and I bet that 95% of your health problems are gone.
The way Tellmegen presents the results is not ideal. You have to click yourself to find the interesting information, filter out what’s not relevant, etc.
For example, this is how I found the relevant information regarding my predisposition to diabetes.
Pros:
Cons:
Diseases are sorted alphabetically
I specifically had to search for type 2 diabetes
You get a nice summary and know exactly what is important.
Pros: - After you find the interesting genes, you can go on and do a deeper research into the topic.
Cons: - You feed your private data to the LLM provider.
Here are for example my results:
| Priority | Gene/Pathway | SNP | Genotype | Association | What It Means |
|---|---|---|---|---|---|
| HIGH | TCF7L2 | rs7903146 | TT | Type 2 Diabetes | ~2x increased risk; strongest diabetes gene |
| HIGH | MTHFR | rs1801133 | AA (TT) | Folate/Homocysteine | 30% enzyme activity; may need more folate |
| HIGH | LCT | rs4988235 | GG | Lactose Intolerance | Likely lactase non-persistent |
| MODERATE | FTO | rs9939609 | AT | Weight/Obesity | One risk allele; modest weight influence |
| MODERATE | COMT | rs4680 | AA | Dopamine | Met/Met; higher prefrontal dopamine |
| MODERATE | ACTN3 | rs1815739 | CC | Muscle Type | α-actinin-3 present; power/sprint advantage |
| MODERATE | CYP1A2 | rs762551 | AA | Caffeine | Fast caffeine metabolizer |
| LOW | SMAD7 | rs4939827 | TT | Colorectal Cancer | Modest risk increase (1.2x) |
| LOW | HNF1B | rs4430796 | AA | Prostate Cancer | Modest risk increase (1.15x) |
You may find the prompt in this GitHub repo.