A San Diego biotech, 43 strong, has spent the last three years teaching small molecules to read cancer's typos. Two of them now do it inside human beings.
Somewhere inside a Phase 1 trial, a drug named ALTA2618 is bumping up against an enzyme called AKT1. It is looking for one specific typo - a glutamic acid where a lysine ought to be. Position 17. When it finds the typo, it binds. When it sees the spelled-correctly version, it walks past.
This is what Alterome Therapeutics built. Not a drug that hits every AKT in the body - those drugs already exist and they make people miserable. A drug that knows the difference between the version of AKT1 a healthy cell uses and the version a tumor cell uses. The difference is one letter of code. The molecule, somehow, can tell.
Down the hall, a second drug, ALTA3263, is doing the same trick against KRAS - the most famous oncogene in cancer biology, and for forty years the most stubborn. It is designed to cover roughly 90% of KRAS mutations, including the workhorses G12V and G12D. Not all KRAS. The wrong ones.
The platform is called Kraken, which is the kind of name a biotech picks when it wants you to know it is not modest. It pulls in proprietary co-crystal data, molecular dynamics, free energy perturbation, quantum chemistry, and machine learning, and uses all of them to design small molecules that fit one specific mutated protein and decline to fit the wild-type one.
This is unusual. Most oncology drugs are blunt - they hit a protein family and let the body sort it out. Alterome's bet is that selectivity is buildable, not lucky. That the difference between "works on the cancer" and "wrecks the patient" is computable, given enough physics and enough data.
They use computers, hard, to design pills that find cancer's broken letters and leave the rest of you alone.
"A KRAS inhibitor designed to hit ~90% of KRAS mutations, including G12V and G12D - the variants that have stayed undruggable longest."
Proprietary co-crystal structures · Molecular dynamics · Free energy perturbation · Quantum chemistry · ML-guided design.
A small molecule designed to inhibit AKT1 only when it carries the E17K mutation - the form found in breast, endometrial, and other cancers. The wild-type AKT, the one your healthy cells need, is meant to walk free.
A KRAS inhibitor designed to cover roughly 90% of KRAS mutations, including G12V and G12D - the most common variants in pancreatic, colorectal, and non-small-cell lung cancer.
Founded in 2021. Series A of $64M led by OrbiMed in 2023. Series B of $132M in April 2024, led by Goldman Sachs Alternatives. The cap table reads like a checklist of life-sciences specialists who have done this before: OrbiMed, Vida Ventures, Nextech Invest, Boxer Capital, Canaan Partners, Invus, Driehaus, Digitalis Ventures, Blue Owl Capital, Colt Ventures.
Josh Richardson of Goldman Sachs Alternatives and Uwe Schoenbeck of Canaan joined the board with the Series B - a signal that the next chapter is about disciplined clinical execution, not platform storytelling.
Co-founder and CEO Eric Murphy, Ph.D. spent roughly twenty years in oncology drug discovery before Alterome, including a stint as co-founder and chief scientific officer at Kinnate Biopharma. Co-founder Ryan Corcoran, M.D., Ph.D. brings clinical and translational oncology depth.
In February 2026, Leo Faoro, MD, MBA joined as Chief Medical Officer to lead the clinical development of both ALTA2618 and ALTA3263 - a hire that says "two trials, both running, both serious."
The company is small - around 43 people - and has been named a Best Place to Work for three consecutive years (2023-2025). It also landed on Fierce Biotech's Fierce 15 in 2025, which is the industry's shorthand for "watch this one."
Co-Founder & CEO. Previously co-founder and CSO at Kinnate Biopharma.
Co-Founder. Clinical and translational oncology.
Chief Medical Officer (2026). Veteran medical oncologist.
Precision oncology has been promising the same thing for fifteen years: medicine that fits your specific cancer, not cancer in general. Most of the time it has delivered drugs for a few star mutations (EGFR, BRAF, the KRAS G12C special case) and shrugged at the rest.
Alterome's pitch, distilled, is that "the rest" is a software problem. Build a good enough computational pipeline, layer it onto real co-crystal structures, and the next mutation becomes designable instead of mysterious. ALTA2618 and ALTA3263 are the first proofs. The pipeline behind them is, in theory, longer.
Three straight years of Best Place to Work recognition is not a small thing in a sector where talent burns fast. The team is intentionally cross-disciplinary - structural biologists, medicinal chemists, computational scientists, oncologists - and small enough that those people have to talk to each other every day. Which is, presumably, the point.
They run on Google Cloud, ship internal comms on Slack, and keep their public face on a WordPress site behind WP Engine - a refreshingly humble tech stack for a company doing quantum chemistry on protein binding pockets.
That same Phase 1 trial. The same drug, the same enzyme, the same typo. What is different is that the trial is now part of a portfolio, not a hope. There is a sibling molecule running its own trial down the hall. There is $200M+ behind both. There is a chief medical officer in the room.
The thing that started as a thesis - that you can design a small molecule to know the difference between healthy biology and broken biology - is now an enrollment number. It is still early. Phase 1 is Phase 1. But the shape of Alterome Therapeutics in 2026 is the shape of a thesis that has bought itself the right to be tested in people. The Kraken has reached out and grabbed two targets. The interesting question is what it grabs next.