Ligo Biosciences

What do they actually do

Ligo Biosciences builds generative deep‑learning models to design new enzymes for industrial chemistry. The team runs an in‑silico design to wet‑lab validation loop: they publish and maintain protein‑modeling code (e.g., an AlphaFold3 implementation) and recruit both ML researchers and wet‑lab scientists with high‑throughput screening experience, indicating they generate enzyme candidates computationally and test them experimentally (website, GitHub).

They are an early YC S24 startup based in San Francisco and appear to be operating in pilot/partner mode rather than offering a broad self‑serve product today. Public materials do not list paying customers; press mentions indicate early collaborations to access sequence data and assay capacity, consistent with building out pilots and infrastructure (YC page, LinkedIn, press).

Who are their target customer(s)

• Industrial chemical manufacturers (bulk and specialty): They need catalysts that lower energy use, cost, and hazardous waste, but existing processes rely on harsh chemical catalysts that are expensive to replace and hard to retrofit into plants (website; press).
• Pharmaceutical process development teams: They face slow or low‑yield steps in small‑molecule synthesis and need selective, scalable enzymes to improve yields, reduce impurities, and cut cycle times (website).
• Food, agriculture, and environmental remediation groups: They need enzymes to break down toxins or contaminants under real‑world conditions but often lack bespoke enzymes that work at scale and under relevant process constraints (website).
• Mid‑sized chemistry/biotech firms without in‑house ML + high‑throughput labs: Building combined computational design and high‑throughput experimental capabilities is costly and slow; they need an external partner to propose sequences and validate them end‑to‑end (website; GitHub).
• Corporate sustainability/decarbonization teams at chemical companies: They are tasked with cutting emissions and energy use in manufacturing but lack proven enzyme replacements that meet throughput, cost, and regulatory constraints for existing processes (website; press).

How would they acquire their first 10, 50, and 100 customers

• First 10: Run tightly scoped paid pilots with industrial R&D teams sourced via YC/warm intros and targeted outbound; convert a computational design into an assayed enzyme using partner foundries or in‑house screening, then publish non‑confidential case studies (YC page; website; press).
• First 50: Package the pilot into repeatable scopes/SOWs, automate handoffs to foundry partners to raise throughput, and use case studies plus conference talks and targeted outreach to win more pilots (website; press).
• First 100: Offer a hybrid product/service (self‑serve design API or subscription for sequence proposals plus managed validation), build channel partnerships with CROs/foundries and consultants, and secure a few enterprise anchors with standardized pricing and SLAs while continuing to publish tooling to maintain credibility (GitHub; website).

What is the rough total addressable market

Top-down context:

Near‑term, the biocatalysis/custom enzyme market is roughly USD ~600–640M in the mid‑2020s; the engineered‑enzymes market is about USD ~2.78B in 2024, and the broader global enzymes market is ~USD 14B (PSMR; FMI; Grand View; IMARC).

Bottom-up calculation:

Illustratively, if pilots are priced at $250k–$750k and the team can run 40–100 pilots per year via partners and in‑house capacity, that implies $10–75M in annual revenue within the biocatalysis SAM; if moving to engineered‑enzyme licensing at $1–5M per program with 10–30 programs/year, that implies $10–150M. These are scenario checks, not forecasts.

Assumptions:

• Average pilot/engagement pricing falls in the $250k–$750k range; licensing programs command $1–5M depending on scope and rights.
• Throughput scales via foundry/in‑house capacity to 40–100 pilots/year or 10–30 licensing programs/year within a few years.
• Focus is on custom biocatalysis and engineered enzymes (not commodity detergent/food enzymes), aligning with the cited market segments.

Who are some of their notable competitors

• Arzeda: Computational protein‑design company that invents new enzymes for industrial uses; overlaps directly on AI‑driven enzyme design and industrial pilots (site).
• Codexis: Established enzyme engineering firm selling bespoke and catalog enzymes; competes on delivering validated, production‑ready biocatalysts to pharma and industrial customers (site).
• Ginkgo Bioworks: Large organism‑engineering foundry offering enzyme engineering and high‑throughput development as a service; competes where customers want end‑to‑end design and scale (site; enzyme services announcement).
• Protera (Protera Bio): AI‑driven protein/enzyme design focused on food and ingredients; overlaps on generative protein design and can win in food/feed applications needing bespoke enzymes (site; industry write‑up).
• Conagen: Synthetic‑biology firm developing strains, enzymes, and fermentation for specialty chemicals and ingredients; competes on applied enzyme solutions with ability to scale to production (site).