The Top Bio and Health Startups in YC W26
YC's W26 batch has one of the deepest concentrations of bio and health companies in recent memory. Here are the ones we're watching most closely ahead of Demo Day.
Ditto Bio
Evolutionary therapies for autoimmune disease
Ditto Bio transforms parasite strategies for modulating the immune system into drug candidates for autoimmune disease. The premise: viruses, ticks, and worms are the natural enemies of the human immune system, and over millions of years they've developed extraordinarily effective molecules to suppress immune response, making their proteins promising drug candidates for diseases where the immune system attacks the body's own tissues.
The team is three PhD scientists from UCSF, UCSD, and UC Berkeley with deep expertise in parasite biology, computational biology, and AI-based protein structure prediction. They worked together for three years as early scientists at another startup before founding Ditto. Since founding seven months ago, they've analyzed over 1M proteins across primate-infecting organisms, discovered thousands of proteins predicted to hit clinically validated human targets, and shown early experimental results with extremely tight binding at 1-2 nM.
CellType
Foundation models and AI agents that simulate human biology to discover drugs
CellType is building what it calls the "agentic drug company," using foundation models and AI agents that simulate human biology to discover drugs. The company's core technology, Cell2Sentence, was co-developed at Yale and published at ICML.
Co-founder David van Dijk is a Yale Professor with 11,000+ citations and publications in Cell, Nature, ICML, and NeurIPS who turned down Google to build CellType. Co-founder Ivan Vrkic co-built the core tech at Yale, previously led large-scale foundation model training at a biotech company, and built software to control CERN's Large Hadron Collider. The company has also released a CLI tool that plans analyses, runs them across 100+ public research tools, and covers target discovery, chemistry, structure, expression, safety, clinical, regulatory, and PK workflows.
Prana
An AI primary care doctor in your pocket
Prana is an always-on AI doctor that replaces the reactive annual physical with continuous, proactive monitoring, connecting to medical records and wearables to detect "clinical drift" like rising blood pressure or glucose weeks before it becomes an emergency. Unlike most AI health apps that are just chatbots, Prana positions itself as a full-stack medical provider whose AI acts as an always-on physician that watches patient data, explains labs, and manages prescriptions.
CEO Meer Patel deferred Brown Medical School to build the company, and CTO Vishvam Rawal comes from quantitative research and algorithmic trading, bringing high-reliability production systems experience to healthcare. Co-founder Sanjit is a graduating MD. The company says it automates 90% of grunt work like history taking, logistics, and triage so its human doctors can focus on high-leverage decisions, enabling 24/7 physician care at a fraction of traditional cost.
Synthetic Sciences
AI co-scientists for end-to-end scientific research
Founded by Aayam Bansal and Ishaan Gangwani, Synthetic Sciences is a platform where scientists delegate complex research tasks to AI co-scientists that handle the full research loop from literature review through hypotheses, experiments, results, and paper drafts. While AI has been showing traction in math research because it's easy to measure, the founders see a larger unlock in accelerating ML research, computational biology, and proteomics where the tooling hasn't caught up.
Synthetic Sciences raised $1.4M pre-YC, and the company is already offering its platform at syntheticsciences.ai.
OctaPulse
Robotics and computer vision for fish farming
OctaPulse is applying robotics and computer vision to aquaculture. The founders saw firsthand the damage being done to the world's oceans, learned that fish is the main protein source for almost 55% of the global population, and that the U.S. imports 90% of its seafood.
OctaPulse uses AI vision to automate hatchery QA for fish farms, starting with broodstock phenotyping and juvenile deformity inspection. The company says it cuts inspection time from about five minutes to under 30 seconds per fish with more than 90% accuracy, so farms advance only high-quality fish and waste less feed and labor.
Mantis
Physics-enhanced synthetic data for low-resource machine learning
Mantis solves one of the biggest bottlenecks in AI: the need for massive amounts of labeled training data. The system uses LLMs and physics simulations to take a small set of real-world observations and automatically generate thousands of realistic, scientifically consistent synthetic training examples. The goal is to unlock high-stakes AI applications that have long been held back by data scarcity, like disease surveillance, rare condition diagnosis, and robotics in novel environments.
Founder Georgia Witchel is a former elite ice climber who helped build the U.S. team, studied CS and psychology at Harvey Mudd, computational math at Johns Hopkins, and biomedical engineering at the University of Washington. Before founding Mantis, she built Louiza Labs, a physics engine powering digital twins for autonomous robotic surgery and simulated FDA trials.