John Lepore, CEO & President - ProFound Therapeutics, on Proteome, AI, & Bold First-in-Class Drugs

What It Covers

John Lepore, CEO of ProFound Therapeutics and CEO Partner at Flagship Pioneering, explains how his company identifies tens of thousands of proteins beyond the original Human Genome Project's 20,000, using this expanded proteome platform to generate first-in-class drug targets across oncology, immunology, and cardiovascular metabolism, with partnerships signed with Pfizer, GSK, and Novartis.

Key Questions Answered

• •Expanded Proteome Opportunity: The Human Genome Project identified roughly 20,000 human proteins, but by removing definitional restrictions and examining ribosomal translation directly in an unbiased way, ProFound Therapeutics has identified tens of thousands of additional proteins. This unexplored protein space represents a systematic path to first-in-class drug targets rather than competing on already-crowded known targets.
• •Partnership Structuring Framework: Before negotiating work plans or legal terms, align explicitly on what each party defines as valuable output from the collaboration. Lepore identifies trust and shared value definition as the two non-negotiable prerequisites, noting that partnerships requiring extensive legal protection to compensate for distrust are a reliable warning sign before signing.
• •Biotech Hiring Archetype: When scaling from 40 to 100 people, prioritize candidates who demonstrate simultaneous patience and impatience — patient enough to let science develop through proper steps, yet impatient enough to respect runway constraints. Rigid adherence to "how it's always been done" is a disqualifying trait; adaptability within a rigorous scientific framework is the target profile.
• •Early-Stage Company Health Signal: A reliable indicator of sound decision-making in early discovery-stage biotechs is the willingness to stop programs. Companies that terminate projects based on data signal science-driven culture over sunk-cost thinking. Separately, headcount growth beyond operational necessity — particularly before clinical milestones — is a warning sign of runway mismanagement in the current funding environment.
• •Agentic AI for Proteomics: ProFound is building an agentic AI system to query its proprietary database of novel proteins against human genetics, expression analysis, mass spectrometry, and known biological pathways. The system draws causal inferences that linear human analysis cannot achieve at scale. This is being built collaboratively between a newly hired CTO and scientific leads, supplemented by Flagship's centralized Pioneering Intelligence function.