Fathom Therapeutics, a New York drug design company, has closed a $47 million Series A financing round. The company uses quantum chemistry and artificial intelligence to study how molecules behave inside living cells, an approach it argues gives a more accurate picture of how a drug will work than most existing methods. Fathom recently changed its name from Atommap Corp. It was co-founded by Huafeng Xu, PhD, Jesús Izaguirre, PhD, and Yujie Wu, PhD, all of whom have backgrounds in computational molecular modeling.
Most drug discovery tools work from static images of proteins. They show what a protein looks like at a single frozen moment, not how it moves and flexes inside the body. That distinction matters. A drug molecule has to interact with a protein as it shifts and changes shape, and if the model used to design it does not account for that motion, the predictions it produces can be unreliable. According to Global Market Insights, the global AI in drug discovery market stood at $3.1 billion in 2025 and is projected to reach $43.9 billion by 2035, a figure that reflects how much the pharmaceutical industry is investing in better computational tools. High clinical failure rates have made the need for those tools hard to ignore.
Sutter Hill Ventures led the round, which was oversubscribed. Other investors include Chemistry, Alexandria Venture Investments, and Empire State Development's NY Ventures.
AI has been applied to drug discovery for years. Most platforms, though, still rely on structural data that captures proteins in a static state. That approach leaves out information about how proteins move over time, and it is precisely that movement that determines whether a drug will bind to its target without triggering unwanted effects elsewhere. Modeling protein dynamics at scale is technically demanding, which is why most tools have not tackled it directly.
Competition in computational drug design has grown considerably over the past two years. Generative AI tools now design novel chemical structures from scratch, and machine learning systems can predict how proteins interact with potential drug candidates. Fathom's bet is that physics-based simulation, rather than statistical pattern-matching, produces predictions that are more grounded and more trustworthy.
"We backed Huafeng because he wanted to rethink how machines understand molecules, combining the best of physics and machine learning to actually move the frontier in computational drug design." — Keith Loebner, Managing Director at Sutter Hill Ventures.
The funding will go toward expanding Fathom's lab-in-the-loop capabilities and growing its internal discovery programs in diseases where treatment options remain limited. The company's lead program has already moved into animal efficacy studies, which is an early but concrete sign of progress toward clinical testing.
Fathom also plans to expand its external partnerships, through which other biotech and pharmaceutical companies use Microcosmos to support their own drug discovery work. Those collaborations serve two purposes: they generate revenue, and they produce data that the company can use to keep improving its platform. Fathom has offices in New York and Boston, both of which are major centers for biotech research and hiring.
The three co-founders have specific and relevant prior experience. Izaguirre previously worked at Roivant Discovery and Silicon Therapeutics. Wu came from Roivant Discovery and Schrödinger. Together, the founding team is also behind the Anton supercomputer and software widely used for predicting how drug molecules bind to proteins. Their collective track record includes advancing 19 drugs to clinical trials, seven of which received FDA approval.
Microcosmos simulates protein motion at the atomic level. That means the platform models the tiny physical movements of individual atoms inside a protein, rather than treating the protein as a fixed object. These simulations are built on quantum mechanical calculations, the branch of physics that describes how atoms and molecules actually behave. The result, according to the company, is a more accurate prediction of how a drug candidate will perform inside a living cell.
"Current drug discovery efforts are limited by reliance on static structures of isolated proteins. Microcosmos is a world model of drugs in living cells that surpasses these limitations by translating accurate quantum mechanical calculations to measurable cellular outcomes." — Huafeng Xu, PhD, Co-Founder and CEO of Fathom Therapeutics.
One result the company has pointed to involves a so-called undruggable target. These are proteins that researchers have long wanted to treat but could not, because their structure offers no obvious place for a drug molecule to attach. Using Microcosmos, Fathom produced potent, highly selective candidates for one such target within six weeks. Those candidates are now being optimized for advancement into the clinic.
Sutter Hill Ventures, founded in 1962, has previously backed NVIDIA, Snowflake, Pure Storage, Corcept Therapeutics, GRAIL, and Forty Seven. Its investments span semiconductors, cloud infrastructure, and life sciences.
"He and the world-class team he's assembled have demonstrated that their approach not only works, but works better than we imagined. We're proud to lead this round, turning Fathom's field-defining molecular design capability into new medicines for patients." — Keith Loebner, Managing Director at Sutter Hill Ventures.
The company also announced two leadership additions. Mandana Honu, PhD, joins as Chief Business Officer, having previously held roles at Protillion Biosciences, Kaleidoscope Bio, and Resilience. Diala Ezzeddine, PhD, joins the board of directors. She has more than 20 years of experience founding and scaling biotech companies, including X-Chem Pharmaceuticals and Magnet Biomedicine. Fathom has additionally formed a Scientific Advisory Board, whose members include Ian Taylor, PhD, former Chief Scientific Officer of Arvinas; Dimitris Agrafiotis, PhD, FRSC, formerly Chief Information Officer at Novartis Institutes for BioMedical Research; and Bruce Zetter, PhD, Professor of Cancer Biology at Harvard University and Boston Children's Hospital.
