Flagship merges three biotechs into Etcelera, an AI bet on knowing which drugs will work before the trial

Flagship Pioneering on Thursday, Oct. 29, 2026, formed Etcelera by merging the platforms of three of its biotech startups, aiming to use AI and "causal biology" to predict how a medicine will act in patients before it reaches the clinic.

Half of drugs that make it to Phase 2, the mid-size human trial, still fall over, and that's the most expensive place to find out you were wrong. Etcelera's bet is that the answer is cause-and-effect biology plus AI, not just more correlation-mining, the hunt for patterns that show up together. It's also, let's be honest, three startups being folded into one with no new money, which is usually what happens when each alone wasn't enough. The rocket line is a great quote; the proof will be the first Etcelera-picked drug that actually survives Phase 2.

Flagship Pioneering announced Etcelera on Oct. 29, 2026. The new company merges the platforms of three Flagship startups: Cellarity, Etiome, and Empress Therapeutics. A platform, here, is the set of tools and data a company uses to find drugs, not a single pill. Those lines are Flagship's, from its press release the same day. Fierce Biotech reports the company is based in Somerville, Mass.

The pitch is what Flagship calls causal intelligence. The job is to figure out what actually drives a disease, then predict how a medicine will behave in a patient before a clinical trial. A clinical trial is the test in people. Flagship says the platform combines human transcriptomic data, chemoproteomics, and large-scale perturbation data with its in-house AI group, Pioneering Intelligence. Transcriptomic data is a readout of which genes are switched on in a cell. Chemoproteomics asks which proteins a chemical sticks to. Perturbation data is the record of what happens when researchers deliberately change a cell and watch. Those lines are Flagship's.

Flagship says the combined platform builds on more than 20 years of data from the three companies. That includes what Flagship calls the industry's largest foundation model, trained on data from 150 million cells, plus more than 30 proprietary AI prediction models. A foundation model is a large model trained on a wide pile of examples, then aimed at many jobs. Proprietary means Flagship owns the models. They are not a public download. The "largest" claim, the cell count, and the model count are Flagship's.

Programs at launch cover hematology, metabolism, and immunology. Hematology is blood disease. The blood programs include myelofibrosis, a disorder that scars the bone marrow, and essential thrombocythemia, too many platelets, the cell fragments that help blood clot. Metabolism includes partnered programs in MASH, short for metabolic dysfunction-associated steatohepatitis, a serious form of fatty liver disease. Immunology is the immune system. Those disease areas are Flagship's. Fierce Biotech reports that Etcelera inherits Cellarity's existing MASH collaboration with Novo.

Cellarity's sickle cell program stays out of the merger. CLY-124 is a clinical-stage small molecule, a chemically made drug already in human trials. Sickle cell disease bends red blood cells into a crescent. CLY-124 stays with a separate company, Cellarity SCD, Inc. That split is Flagship's.

Ad Rawcliffe is chief executive. He previously led Adaptimmune, and he is a Flagship CEO-Partner who led Etiome. Flagship senior partner Alec Reynolds is chairman. Flagship chief executive Noubar Afeyan is a co-founder. CEO-Partner is Flagship's title for an executive it places inside one of its companies. Those roles are Flagship's.

BioPharma Dive reports that no new funding was announced with the merger. The three companies had raised $374 million in venture capital before they were combined. Venture capital is money from investors who take a stake in a young company. That funding line is BioPharma Dive's.

Alec Reynolds said, "Approximately 50 percent of drug candidates that reach Phase 2 fail because the industry lacks causal intelligence." Phase 2 is the mid-size human trial, the expensive step where a company tries to show a drug actually works. That quotation is his, in Flagship's release.

Ad Rawcliffe said, "We can predict where a rocket will land, yet we still can't predict what a medicine will do in a patient." That quotation is his, in the release.

The picture is a portrait of Etcelera chief executive Ad Rawcliffe. The credit is Flagship Pioneering, via C&EN.

In plain terms, Flagship on Thursday folded three of its biotech startups into one company, Etcelera, and put Ad Rawcliffe in charge. The bet is an AI system, trained in part on 150 million cells, that tries to say how a drug will act before a human trial. The first programs are in blood disease, fatty liver disease, and the immune system. A sickle-cell drug stays in a separate company. No new money was announced. The three startups had already raised $374 million.