About Fabricagen
We build the infrastructure drug discovery was missing.
Drug discovery has run the same design-test loop for thirty years. AI tools now generate protein candidates in minutes. Experimental validation still takes months. The two systems have never communicated — the model never learns from what the lab finds.
Fabricagen closes the loop. Our platform combines a generative diffusion model with a proprietary functional display assay and in vivo pharmacology, integrated into a one-week design-test-retrain cycle. Every experiment produces data that improves the next design round. The platform compounds with every campaign it runs.
We work with pharmaceutical and biotechnology companies who need validated candidates faster than legacy timelines allow. We also build our own pipeline from the discoveries the platform makes.
01
The moat is the loop
The generative model and the display assay were designed together, for each other. Neither is novel without the other. Per-candidate functional labels flow directly into model retraining — no competitor has this combination.
02
Data compounds
Every campaign produces target-specific sequence-function data that no one else has. The model learns from cell assay results, conjugate characterization, and in vivo pharmacology. The platform gets smarter with use.
03
Lab + compute, not either/or
Most protein design companies are software only. Most CROs are wet lab only. Fabricagen owns both ends of the loop because closing it required expertise in protein physics, display chemistry, and machine learning simultaneously.
Team
Michael R. Holden, PhD
Founder & CEO
Michael built both the model and the assay. The combination of protein physics, wet-lab throughput, and machine learning is what made the closed-loop design possible.
He received his PhD in Biochemistry from the University of Denver (2018), studying tau aggregation kinetics and single-molecule biophysics. His postdoctoral research at the University of Colorado School of Medicine focused on the structural biology of chromatin-modifying enzymes.
In industry: at Prothena Biosciences, he contributed to antibody discovery programs across tau, alpha-synuclein, and TDP-43 — preparing fibril test articles from human brain tissue that were used to develop the drug candidates. At Regeneron as Principal Scientist and Research Group Leader, he led an antibody engineering team. At BioLoomics as Head of Chemistry, he built the company's entire antibody-to-ADC workflow, producing more than 2,500 antibodies and 1,000 drug conjugates in one year using novel site-specific conjugation chemistry.
Eight peer-reviewed publications. Visiting Scholar, University of Denver.
Mike Minson, PhD
Founding Scientist
Mike runs the wet-lab arm. Cellular biochemist and molecular biologist with seven years of in vitro assay development across diagnostics, biologics, and synthetic biology.
He received his PhD from the University of Colorado Boulder in 2019. At ArcherDx, he led NGS assay design and validation work supporting an FDA submission. At Sartorius, antibody and exosome platform work. At BioLoomics, he was a senior scientist and synthetic-biology lead — he built the platform that supported the company's $8M raise and ran the in vitro assay strategy across the ADC discovery campaigns.
Mike also designs binders computationally and validates them at the bench: his bispecific binder against RBX1 placed in the GEM-X / Adaptyv RBx1 Binder Design Competition — one of nine confirmed binders out of 322 designs (180 nM Kd).
Holden and Minson worked together at BioLoomics on ADC discovery before Mike joined Fabricagen as Founding Scientist in May 2026.