Ligo Biosciences

Ligo Biosciences develops computational tooling for chemistry and biology, beginning with enzyme design. The company is building foundational deep-learning models that learn the principles of catalysis and generate protein structures capable of catalysing a target reaction directly from transition-state models, rather than starting from an enzyme that already exists. The stated aim is to widen the set of reactions that can feasibly be accelerated by enzymes, including the synthesis of high-value chemicals used in pharmaceuticals, agriculture and consumer goods.

The company frames its work around two premises: that general-purpose intelligence such as large language models is becoming abundant and able to reason, write code, analyse data and design experiments across scientific domains; and that specialised computational tools are the limiting factor for scientific progress in domains that general reasoning cannot address. It cites AlphaFold as the reference case, where a neural network trained on the Protein Data Bank created a computational interface to protein geometry that both humans and software agents could then reason over. Ligo positions enzymes as its first domain because they combine large potential impact with fast experimental feedback loops.

In its Y Combinator launch material the company describes the chemical industry as a roughly $6 trillion market that accounts for 20% of industrial greenhouse gas emissions and 15% of global energy use, and notes that enzyme development for manufacturing has historically cost tens of millions of dollars and taken multiple years. Ligo's positioning is that its models can compress enzyme design timelines from years to days. The team is described as a small group of researchers spanning biology, chemistry, mathematics and computer science.

Founding story

Ligo Biosciences was founded in 2024 by Edward Harris (CEO), Emily Egerton-Warburton (CSO) and Arda Goreci (CTO), who met in a synthetic biology lab at Oxford University and decided to target the chemicals industry. Harris studied computer science at Princeton before transferring to Oxford Medical School and working across three synthetic biology labs; he had earlier bootstrapped a first company at 19 in the food markets of Guadalajara, Mexico, which he reports reaching $1M in annual revenue. Egerton-Warburton is a biochemist from Oxford with wet-lab experience at biotech startups, having worked on projects ranging from bacterial biofuel production to vaccine design. Goreci studied cell and systems biology at Oxford and was named a Google Cloud Research Innovator for work in computational biology, with an interest in deep learning, scaling laws, geometric deep learning and biomolecular design dating to the original AlphaFold paper.

Business model

Ligo builds proprietary enzyme design models and offers designed enzymes and design capability for reactions needed by chemical and life-science manufacturers, with a waitlist used for early access. Alongside this it has published open-source model work. Public sources do not detail pricing or contract structures.

Not specified in available sources.

Traction

Y Combinator Summer 2024 participation with an active status and a team of four. Announced partnerships with Basecamp Research, described as on track to hold 1000x more sequence diversity than public resources, and with Adaptyv Bio, whose engineers were helping build the data pipeline for Ligo's open-source model using the ProteinFlow bioinformatics tool. An open-source implementation of AlphaFold3 was published in September 2024. A waitlist was open for the forthcoming open-source model.

Latest developments

Founders Arda Goreci and Edward Harris were featured in a May 2025 episode of the Sustainable Times "Profit Meets Purpose" podcast. The company website's current positioning describes Ligo as building next-generation tooling to manipulate chemistry and biology, starting with enzyme design, and frames its work around enabling both researchers and autonomous scientific agents to search enzyme space.