AI DNA Analysis Is Going Viral — But Should You Send Your Genome to the Cloud?

In February 2026, entrepreneur and AI automation expert Nick Saraev shared a LinkedIn post that quickly went viral, racking up thousands of reactions and comments within days of posting. His claim: he uploaded his 23andMe raw DNA data to Claude Code (Anthropic's AI coding assistant, which sends file contents to Claude's API for cloud-based processing), built an analysis pipeline that cross-referenced his genetic data against medical databases, and the results “genuinely changed my life.” The specific discoveries he reported were striking:

• Cystic fibrosis carrier status — he discovered he carries a gene for CF, which has real implications for family planning
• Methylation pathway dysfunction — the analysis flagged issues with his methylation pathways and recommended methylfolate supplementation. He reported dramatically improved energy and sleep “within 2 to 3 days”
• Poor caffeine metabolism — his CYP enzyme variants showed he metabolizes caffeine slowly, which was disrupting his sleep
• Personalized health protocol — dietary frameworks, exercise protocols, and supplement recommendations tailored to his genetic profile

His framing resonated powerfully: “This kind of personalized genetic health analysis used to cost $5,000 to $10,000. I got my DNA tested through 23andMe for about $70 Canadian.” The post struck a nerve because the transformation was tangible and specific. Not vague promises about “the future of health” — real changes to sleep, energy, and daily routine that he could feel within days.

Nick's core message is correct, and it's one we believe in deeply: the raw DNA data sitting in your 23andMe or AncestryDNA account contains actionable health information that most people never access. Carrier status is life-changing information. Discovering you're a carrier for conditions like cystic fibrosis, sickle cell disease, or Tay-Sachs disease has direct implications for family planning. If both parents carry the same recessive gene, their children have a 25% chance of inheriting the condition. This is information worth having before starting a family. MTHFR variants are extremely common and actionable. The MTHFR gene produces an enzyme critical for folate metabolism. Common variants like C677T and A1298C can reduce enzyme efficiency, potentially affecting folate metabolism, homocysteine levels, and cardiovascular health. Some research has explored connections to energy and mood, though evidence for direct causal effects remains mixed. Switching from standard folic acid to methylfolate is an approach discussed in the research literature for people with reduced MTHFR function, though evidence for widespread clinical benefit remains an active area of study. Some people with reduced MTHFR function report improvements after switching from folic acid to methylfolate. Nick's reported experience may reflect this, though individual responses vary and self-reported outcomes should be interpreted cautiously. Discuss any supplement changes with a healthcare provider. Caffeine metabolism varies dramatically by genotype. Variants in CYP1A2 determine whether you're a fast or slow caffeine metabolizer. Slow metabolizers who drink coffee late in the day (or too much of it) can experience disrupted sleep, anxiety, and elevated blood pressure — often without realizing caffeine is the cause. Learning your caffeine metabolism status is one of the most immediately actionable pharmacogenomic insights available from consumer DNA data. None of this is speculative. These are well-characterized genetic variants with published research supporting their clinical relevance. The databases Nick's pipeline used — ClinVar and PharmGKB — are the same peer-reviewed, NIH-supported resources that underpin clinical genomics.

Here's where it gets complicated. To build his analysis pipeline, Nick uploaded his raw DNA file to Claude Code, which processes commands through Anthropic's cloud servers. Several commenters on the post raised privacy concerns directly, questioning whether sharing an entire gene profile with a cloud-based AI service was prudent. Let's be clear about what's at stake:

• Your genome never changes. A password can be reset. A credit card can be canceled. Your DNA is permanent and uniquely identifies you for life.
• Your DNA reveals information about your family. Your genetic data contains information about your parents, siblings, children, and extended relatives — people who never consented to having their genetic patterns shared.
• Cloud-processed data creates a permanent record. Once your genetic data has been transmitted to external servers, you lose control over how it's stored, how long it's retained, who can access it, and whether it could be exposed in a future data breach.
• AI model providers have their own data policies. While many AI companies state they don't train on user data, terms of service can change, companies can be acquired, and server logs may persist even after data is “deleted.”

This isn't a criticism of Nick or his approach — he's a technical user who made an informed choice, and he explicitly noted he's not a doctor. But the thousands of people inspired by his post to do the same thing may not fully understand what they're giving up when they paste 600,000 to 700,000 lines of genetic data into a cloud-based AI tool. This is not an abstract concern. The 23andMe data breach exposed the genetic information of approximately 6.9 million people — demonstrating that server-side storage of genetic data carries real, documented risk. Note: 23andMe filed for bankruptcy in early 2025 and was acquired by TTAM Research Institute in July 2025 — download your data sooner rather than later. That experience is exactly why DNA Explore was built with a fundamentally different architecture.

Nick's approach required Claude Code (Anthropic's AI coding assistant, which sends file contents to Claude's API for cloud-based processing), Python scripting knowledge, manually downloading medical databases like ClinVar (289MB+), and building a custom analysis pipeline from scratch. It's impressive engineering — but it's not realistic for most people. The health insights he discovered aren't locked behind a coding terminal. They're in your raw DNA data file right now, waiting to be read. The same variants that flagged his MTHFR dysfunction, caffeine metabolism, and carrier status are sitting in your 23andMe or AncestryDNA raw data file. You don't need to write code to access them. DNA Explore was built to solve exactly this problem: make genetic health analysis accessible to everyone, not just developers. You drop your raw data file into your browser. The analysis runs locally on your device. You get results in seconds. Your data never leaves your computer. Here's what DNA Explore covers that maps directly to Nick's discoveries:

• MTHFR and methylation analysis — C677T, A1298C, and related variants with plain-English explanations of what your genotype means for folate metabolism
• Pharmacogenomics (95+ drug interactions) — CYP enzyme profiling including caffeine metabolism (CYP1A2), antidepressant response (CYP2D6, CYP2C19), statin risk, and more
• Carrier status — analysis of variants associated with inherited conditions
• Polygenic risk scores — multi-condition risk assessment not described in Nick's post
• Nutrigenomics — how your genes affect vitamin needs, food sensitivities, and nutrient metabolism
• Gene-gene interactions — how variants interact with each other to compound or offset risk
• AI-powered explanations — ask questions about your results in plain English

Note: The Claude Code pipeline column reflects what was described in Nick's LinkedIn post. His actual implementation may include additional capabilities not mentioned in the post. Claude Code sends file context to Anthropic's API for processing; always review current terms before uploading sensitive data to any AI service.

For a detailed breakdown of how DNA Explore compares to other tools, see our full comparison page.

Inspired by Nick's experience, here are five categories of genetic insight that are already sitting in your raw data file — and that DNA Explore analyzes automatically: 1. Carrier status for inherited conditions You could be a silent carrier for conditions like cystic fibrosis, sickle cell disease, hereditary hemochromatosis, or dozens of other recessive conditions without ever knowing it. If your partner carries the same gene, your children could inherit the condition. This is information that can genuinely change the trajectory of a family. 2. How your body processes medications Your CYP enzyme variants determine how you metabolize everything from antidepressants to pain medications to caffeine. Widely cited research from the Clinical Pharmacogenetics Implementation Consortium suggests roughly 90% of people carry at least one actionable pharmacogenomic variant — though exact figures vary by how broadly interactions are defined. If a medication “doesn't work” for you or gives you unusual side effects, your genes may be why. 3. MTHFR and methylation The MTHFR gene variants C677T and A1298C are carried by a substantial portion of the population. Reduced MTHFR function can affect folate metabolism, which is involved in energy production, mood regulation, and cardiovascular health. For some people, switching from standard folic acid to methylfolate can make a noticeable difference — as Nick reported experiencing. 4. Disease predisposition through polygenic risk scores Polygenic risk scores aggregate the effects of hundreds of genetic variants to estimate your predisposition to conditions like coronary artery disease, type 2 diabetes, and certain cancers. Research suggests that people in the top 5% of genetic risk may face significantly elevated risk — sometimes comparable to rare high-impact mutations. This isn't a diagnosis, but it can motivate earlier screening conversations with your doctor. 5. Nutrient metabolism and dietary needs Your genes influence how you process vitamins, fats, carbohydrates, and other nutrients. Variants in genes like FUT2 affect vitamin B12 absorption. LCT variants determine lactose tolerance. FADS1/FADS2 variants affect omega-3 and omega-6 metabolism. This information can help you customize your diet to your actual biology rather than following generic advice.

If Nick's post inspired you to explore your own genetic data, here's how to do it without sending your genome to the cloud:

1. Get your raw data file. If you've already taken a 23andMe or AncestryDNA test, your raw data is free to download from your account. (Note: 23andMe filed for bankruptcy in early 2025 and was acquired by TTAM Research Institute in July 2025 — download your data sooner rather than later.) If you haven't tested yet, a consumer kit costs about $79 on sale.
2. Choose a privacy-first analysis tool. Look for tools that process your data locally in your browser — not on external servers. DNA Explore is built on this principle: your raw data file never leaves your device.
3. Review your results with context. Pay attention to your MTHFR status, CYP enzyme variants (especially caffeine and drug metabolism), carrier status, and polygenic risk scores.
4. Talk to your doctor. Bring your results to a healthcare provider who can contextualize them alongside your medical history, family history, and current health. Genetic insights are a starting point for conversation, not a self-prescribing tool.

The discoveries Nick made — MTHFR dysfunction, caffeine metabolism, carrier status — are exactly the kind of insights that DNA Explore surfaces automatically from your raw data, without requiring a terminal, a coding background, or sending your genome to any server.

I'm Peter Hollens — musician, entrepreneur, and one of the millions of people who had their data exposed in the 23andMe data breach. After that breach, I wanted to understand what my DNA said about my health. But every tool I found required uploading my genetic data to someone else's server. After having my data compromised once, I wasn't about to hand it over again — not to a startup, not to an AI model, not to anyone. So I built DNA Explore with one non-negotiable principle: your genetic data never leaves your device. Not as a privacy policy that could change with a terms-of-service update. As an architectural guarantee. The analysis runs entirely in JavaScript in your browser. There's no server to breach, no database to hack, and no upload to intercept — because the data never goes anywhere. Nick Saraev's viral post is a powerful demonstration that your DNA contains life-changing information. We agree completely. We just believe you shouldn't have to sacrifice your privacy to access it.