The money came in quietly, then all at once.
Transcripta Bio, the Palo Alto drug discovery company wagering that artificial intelligence can crack neurological diseases—including rare conditions often deprioritized by Big Pharma—closed $24 million in post-Series A capital, a sum the startup says will carry its experimental therapies for autism spectrum disorder and a rare form of muscular dystrophy toward human trials.
The financing, disclosed in July 2026, pulled in Mayo Clinic and Dallas-based Omnimed Capital as fresh backers. They joined earlier investors JAZZ Venture Partners, BlueYard Capital, and a handful of life sciences family offices whose names the company declined to share. Transcripta isn't saying what the round valued the business at, and the full terms remain private.
For a company built on the premise that machine learning can compress timelines in rare disease development—where patient populations are often too small and fragmented to justify traditional pharmaceutical economics—the new capital marks something closer to a validation. Perhaps more than the founders expected when they rebranded from the obscure "Rarebase" barely a year earlier.
Following the Money (and the Discrepancies)
Tracking Transcripta's fundraising history requires a bit of detective work. JAZZ Venture Partners led an earlier Series A, but public records tell different stories about the size. CB Insights and BlueYard Capital's own portfolio listing both reference a $15 million Series A, also led by JAZZ. An April 2025 blog post from Transcripta mentioned pulling in another $10 million from existing backers. Now the company says it's raised $24 million total since that initial Series A.
The numbers don't quite reconcile cleanly, though this kind of opacity is hardly unusual for private startups navigating multiple tranches and rolling closes. What's clear: Transcripta has cobbled together enough runway to advance four drug programs, two of which—autism spectrum disorder and facioscapulohumeral muscular dystrophy (FSHD)—are now heading toward the regulatory hurdles required before human testing can begin.
Mayo Clinic's participation adds institutional credibility that venture dollars alone can't buy. One of the world's preeminent research hospitals doesn't typically write checks without extensive vetting, and its involvement signals a degree of scientific confidence in Transcripta's approach.
A Billion Genes, Virtually Screened
At the heart of Transcripta's pitch sits Conductor AI, a platform the company claims has ingested data on more than one billion gene responses. The architecture stitches together three components: a Disease Signature Atlas mapping how illnesses alter cellular behavior, a Drug-Gene Atlas cataloging what compounds do to those same cells, and the AI layer connecting the two through "transcriptomic signature matching"—essentially, finding molecular patterns that suggest a drug might reverse a disease's genetic fingerprint.
It's an audacious technical undertaking. A 2024 Nature feature on the company reported that Transcripta had compiled information from over 200 million experiments, mapped thousands of compounds and their effects on 20,000 genes, and trained 16,000 separate machine learning models—one for each gene. The platform demonstrated it could virtually screen 6.5 billion drug-like molecules "in a couple of days," with roughly 20 percent of the 21 synthesized predictions showing cellular activity in lab tests.
That hit rate, if it holds, would represent a meaningful improvement over traditional high-throughput screening methods. But virtual predictions and wet-lab validation are very different animals, and the real test will come when these compounds move into patients.
Transcripta has been accumulating partnerships to bolster its dataset. In May 2025, the company announced a collaboration with Microsoft Research to analyze chemotranscriptomic data for disease-gene links. Around March of this year, it contributed a high-resolution CRISPRi dataset to Verily's Pre Exchange and Workbench platform, a move that expanded its presence in the broader computational biology ecosystem.
Four Programs, No Trials Yet
The fresh capital will fund IND-enabling studies—the preclinical work required before submitting an investigational new drug application to regulators—across Transcripta's pipeline. The company is prioritizing clinical preparation for its autism spectrum disorder and FSHD programs, while also working on earlier-stage efforts in myotonic dystrophy and Huntington's disease.
As of July, none of these programs have entered clinical trials. No studies appear in public registries, and the company hasn't disclosed timelines for when it expects to dose its first human subjects. That's not unusual for a preclinical-stage biotech, but it does mean Transcripta remains years away from proving its platform can deliver approvable therapies—the only metric that ultimately matters.
The company was founded in 2020 under the name Rarebase, a moniker that never quite stuck. It rebranded as Transcripta in April 2025 under CEO and co-founder Chris Moxham, who doubles as chief scientific officer. TIME named the company to its 100 Most Influential Companies list in May 2024, a PR win that likely helped grease the wheels for subsequent fundraising.
The AI Biotech Bet Continues
Transcripta's funding arrives amid sustained, if somewhat indiscriminate, investor enthusiasm for AI-driven drug discovery. Competitor Xaira pulled in roughly $1 billion in April 2024, a staggering sum that reflected both the promise and hype surrounding machine learning in life sciences. The thesis: AI can identify drug candidates faster and cheaper than traditional methods, compressing development timelines in therapeutic areas where patient need far outpaces available treatments.
Whether that thesis holds—and whether platforms like Conductor AI can navigate the notorious valley of death between preclinical promise and clinical proof—remains an open question. For now, Transcripta has bought itself more time to find out.
The company declined to comment beyond its prepared announcement.
