Computational Biology Market

Certara sold USD 183.3 million of modelling software in 2025 and USD 235.6 million of modelling services, yet software grew 18% while services grew 3%, and that gap is the clearest signal of where the computational biology market is heading. Computational biology is the use of simulation, mathematical models and machine learning to predict how molecules, cells, organs and patients behave; the market covers modelling software, model-based contract services and in-house modelling teams’ paid tools, from molecular dynamics to physiologically based pharmacokinetic (PBPK) and quantitative systems pharmacology (QSP) models. Douglas Insights sizes the computational biology market at USD 6.84 billion in 2025, which is 118,600 paid modelling seats multiplied by an average annual spend of USD 57,700 per seat, and expects USD 20.8 billion by 2035, a revenue CAGR of 11.75%. The policy push is dated: on 10 April 2025 the Food and Drug Administration (FDA) announced a plan to phase out animal testing requirements for monoclonal antibodies and named AI-based computational models of toxicity among its new approach methodologies (NAMs). The study belongs to our life science instruments coverage, and every input is listed in the research methodology.

Which application segment of computational biology makes the money, from cellular simulation to computational toxicology?

Cellular and biological simulation leads computational biology with 31.4% of 2025 revenue, or USD 2.15 billion, because every target, pathway and dose model starts from a simulated cell or tissue. Drug discovery and disease modelling follows at 27.9%, while computational toxicology grows fastest at 14.6% a year through 2035.

Cellular and biological simulation is worth USD 2.15 billion in 2025 and reaches USD 5.68 billion by 2035 at 10.2% a year; its share is large because molecular dynamics, pathway and organ models run in every discovery group, academic or commercial. Drug discovery and disease modelling holds USD 1.91 billion and grows 12.7% a year, since physics-based design and AI structure prediction now screen far more molecules than a wet lab can make.

Preclinical drug development and computational toxicology is the smallest of the regulated segments today at USD 1.15 billion, but it is the fastest at 14.6%, reaching USD 4.49 billion by 2035. The reason is regulatory: the FDA roadmap tells sponsors to include NAMs data in investigational new drug (IND) applications, and in silico toxicity prediction is the cheapest NAM to scale.

Clinical trial modelling and simulation carries USD 978.6 million, growing 9.6% a year, because model-informed dose selection and virtual control arms are already routine at large sponsors. Computational genomics is USD 656.9 million at 9.6% share and grows 10.9%; it stays small here because sequencing pipelines and variant interpretation sit in the separate Bioinformatics Technologies Market report.

Why are pharma and biotech teams buying more computational biology modelling seats?

Paid computational biology seats grow 8.6% a year to about 270,634 by 2035, driven by four forces: FDA support for non-animal evidence, AI structure prediction, model-informed clinical development and the volume of omics data. Douglas Insights splits that 8.6-point volume leg into 3.1, 2.8, 1.6 and 1.1 points.

Regulatory acceptance adds 3.1 points. The FDA plan of 10 April 2025 said implementation would begin immediately for IND applications and that a pilot for selected monoclonal antibody developers would follow within a year. On 20 April 2026 the agency reported it had met its first-year goals, qualified its first AI-based drug development tool relying on in silico models, and issued draft guidance widening weight-of-evidence approaches that include computational toxicology. The same release notes that over 90% of drugs that pass animal studies are not approved, which is the commercial case for computational biology in one number.

AI-guided molecular design adds 2.8 points. Schrödinger reported 2025 drug discovery revenue of USD 56.4 million, more than double the USD 27.2 million of 2024, alongside USD 199.5 million of software revenue, according to its fourth-quarter results. Drug discovery and disease modelling seats grow fastest in large pharma, where each free energy perturbation or docking campaign replaces weeks of synthesis.

Model-informed clinical development adds 1.6 points. Certara serves more than 2,600 biopharmaceutical companies, academic institutions and regulatory agencies in 70 countries, and a sponsor that runs one PBPK model for a drug interaction question tends to run one for every later asset. That repeat pattern is why Douglas Insights counts clinical trial modelling seats rising by about 9.6% a year in value terms.

Omics data volume adds the last 1.1 points. Recursion reported 2025 revenue of USD 74.7 million and cumulative partner payments of USD 213 million from Roche and Genentech, built on phenomic maps that only computational biology tools can read; its full-year release also lists USD 134 million in Sanofi milestones to date. The four contributions sum to 8.6 points: 3.1 plus 2.8 plus 1.6 plus 1.1.

Each driver lands in a different computational biology segment. Regulatory acceptance feeds computational toxicology, AI design feeds discovery modelling, and model-informed development feeds trial simulation. Douglas Insights expects the regulatory driver alone to add about 54,803 seats by 2035, the largest single source of new demand.

Which barriers slow computational biology adoption in wet-lab-first research groups?

Three barriers remove about 2.9 points from computational biology volume growth: flat sponsor budgets, model credibility questions and a thin pool of modellers. Without them seats would grow 11.5% a year rather than 8.6%, so the drag is worth about USD 6.27 billion of 2035 revenue.

Budget caution removes 1.2 points. Certara guided 2026 revenue growth of 0% to 4%, and Simulations Plus also guided total revenue growth of 0% to 4% for fiscal 2026 in its fiscal 2025 results, both well below their 2025 software growth. When biotech funding tightens, contract modelling projects are the first line deferred.

Model credibility removes 0.9 points. A computational biology model only displaces an animal or clinical study once a reviewer accepts its context of use, and the FDA pilot for monoclonal antibodies was still a pilot in its first year. Sponsors therefore pay twice, running the model and the legacy study side by side.

Talent supply removes 0.8 points. Physics-based modelling, pharmacometrics and machine learning rarely sit in one person, so a modelling seat often waits months for a qualified user. Douglas Insights treats that as a seat delay rather than lost demand, which is why the slower scenario still grows.

How much does a computational biology modelling seat cost per year, from academic licences to enterprise suites?

A computational biology modelling seat averages USD 57,700 a year in 2025, rising 2.9% a year to about USD 76,794 by 2035. That average blends discounted academic licences, enterprise software and the service hours each seat consumes.

Certara’s 2025 software revenue of USD 183.3 million across more than 2,600 customers works out to about USD 70,500 of software per customer, a useful upper anchor for a commercial seat. Schrödinger reported 27 customers with annual contract value above USD 1 million in 2025, down from 29 in 2024, which marks the enterprise ceiling where a single pharma group buys hundreds of seats and cloud compute together.

Douglas Insights estimates three realised bands: academic and government seats at roughly USD 12,000 to USD 32,000 a year, commercial single-discipline seats at USD 35,000 to USD 75,000, and enterprise seats bundled with contract modelling at USD 90,000 to USD 160,000. BIOVIA Discovery Studio from Dassault Systemes shows the low end of the access curve: its visualiser is free, while simulations need a paid licence, sold as software as a service or on premise.

The 2.9% price leg is modest because cloud delivery lowers entry prices even as advanced modules, such as free energy calculations and biotherapeutics developability, lift spend per seat.

Which companies win computational biology software and modelling services contracts?

Certara leads computational biology with an estimated 6.1% share of 2025 revenue, based on its disclosed USD 418.8 million, and the top three suppliers hold about 11.0% together. The field is fragmented because modelling splits by discipline, from molecules to patients.

Certara grew 2025 revenue 9% to USD 418.8 million, with software up 18% and services up 3%, according to its full-year release of 26 February 2026. Its position rests on regulatory credibility: agencies themselves are among its customers. Schrödinger’s USD 255.9 million total, up 23.3%, comes from licensing the same physics engine it uses in its own pipeline.

Simulations Plus booked USD 79.2 million in fiscal 2025, up 13%, with software at 58% of revenue and services at 42%. Recursion pairs computational biology with partner economics, and its cash of USD 753.9 million at the end of 2025 funds operations into early 2028. Dassault Systemes sells BIOVIA Discovery Studio, which draws on more than 30 years of peer-reviewed methods. Chemical Computing Group, founded in 1994 in Montreal, sells MOE to discovery chemists and runs offices in five countries.

Douglas Insights does not publish shares for BIOVIA or Chemical Computing Group, because neither discloses computational biology revenue. Adjacent protein work sits in the Protein Engineering Market study.

Where does computational biology spending grow fastest, North America or Asia Pacific?

North America leads computational biology with USD 3.17 billion in 2025, a 46.3% share, while Asia Pacific grows fastest at 13.8% a year to USD 5.16 billion by 2035. The split follows where pharma R&D budgets and regulators that accept model evidence sit.

North America holds USD 3.17 billion and grows 11.3% a year, because the FDA roadmap applies first to US IND filings and most listed modelling vendors report there. Europe carries USD 1.79 billion, growing 10.8%, on the strength of large Swiss, German and British sponsors and their contract research partners.

Asia Pacific is worth USD 1.42 billion in 2025 and is the fastest region at 13.8% a year, as Chinese, Japanese and Korean biotechs adopt AI-led discovery without legacy wet-lab capacity. Latin America adds USD 266.9 million, growing 11.9%, mostly through academic and generic-drug modelling.

The Middle East and Africa is the wildcard at USD 205.3 million and 10.91% a year: sovereign genomics and drug discovery programmes in the Gulf could lift it far faster than its 3.0% share implies.

What if FDA new approach methodologies move faster: how do computational biology scenarios to 2035 compare?

The base case reaches USD 20.8 billion in 2035 on 8.6% seat growth and a 2.9% price leg, while slower and faster cases bracket it at USD 16.4 billion and USD 25.9 billion. The swing depends mainly on how quickly regulators accept model evidence.

The slower case uses 6.6% seat growth and a 2.4% price leg, giving USD 16.4 billion; it assumes the FDA pilot that began after 10 April 2025 stays narrow and budgets stay flat. The faster case uses 10.6% seat growth and 3.3% pricing, giving USD 25.9 billion, if computational toxicology replaces whole animal study packages.

Sensitivity is steep. A 1-point shift in annual seat growth changes computational biology revenue in 2035 by about USD 1.92 billion in either direction, so the volume assumption matters more than price. Published forecasts run from about 12.3% to 17.1% a year; our 11.75% sits just below that band because sequencing pipelines are excluded.

Sources

1. US Food and Drug Administration FDA plan to phase out animal testing requirement for monoclonal antibodies (2025)
2. US Food and Drug Administration FDA achieves Year 1 goals in reducing animal testing (2026)
3. Certara Certara fourth quarter 2025 financial results (2026)
4. Schrödinger Schrödinger fourth quarter and full-year 2025 results (2026)
5. SEC EDGAR Simulations Plus fiscal 2025 results, Form 8-K exhibit (2025)
6. SEC EDGAR Recursion full year 2025 results, Form 8-K exhibit (2026)
7. Dassault Systemes BIOVIA Discovery Studio (2026)
8. Chemical Computing Group About Chemical Computing Group (2026)

How does FDA policy on new approach methodologies change the evidence computational biology must produce?

FDA policy now invites computational biology evidence in IND applications: the 10 April 2025 roadmap made NAMs data acceptable immediately, and the April 2026 update added draft guidance on computational toxicology. Model qualification, not software, becomes the scarce output.

The draft guidance on reducing nonhuman primate testing for monoclonal antibodies and the guidance on weight-of-evidence approaches both name in silico methods as admissible inputs. The FDA also reported qualifying its first AI-based drug development tool that relies on in silico models. For computational biology vendors, the compliance task is documentation: context of use, validation data and version control for every model a sponsor submits.

Where does computational biology end and bioinformatics or contract research begin?

Computational biology in this study covers predictive modelling of molecules, cells and patients worth USD 6.84 billion in 2025; it excludes sequencing pipelines, lab notebooks and data storage. That boundary keeps the market distinct from bioinformatics data processing.

The offering split is software platforms, contract modelling services and in-house modelling services. Software platforms carry about 47.6% of value, contract modelling services 38.9% and in-house modelling services 13.5%, a Douglas Insights estimate anchored on Certara’s 44% software mix and Simulations Plus’s 58%. End users are pharmaceutical and biotechnology companies, academic and research institutes, and regulators and contract research organisations; outsourced trial work beyond modelling is sized in the Healthcare Contract Research Organization Market report.

Which end users fund computational biology seats: pharma groups, universities or regulators?

Pharmaceutical and biotechnology companies fund 61.2% of computational biology revenue in 2025, about USD 4.19 billion, because modelling decisions sit directly on their development budgets. Academic and research institutes add 27.4% and regulators and contract research organisations 11.4%.

Douglas Insights puts commercial pharma and biotech spend at USD 4.19 billion, growing faster than the market at roughly 12.4% a year as sponsors move modelling earlier in discovery. Academic and research institutes account for USD 1.88 billion; their seat count is high but their discounted licences keep value per seat near the bottom of the price bands. Regulators and contract research organisations hold USD 780.1 million, and that group counts Certara’s agency customers among its buyers.

Our model credits academia with about 52% of seats but only 27.4% of value, which is the clearest illustration of how computational biology pricing tiers work.

Douglas Exclusive: the computational biology in-silico evidence register

The register is a Douglas Insights model, not an official list. It is built from 14 sourced inputs: six company disclosures, two FDA releases, two FDA guidance items named in those releases and four published growth estimates. It maps where computational biology evidence already substitutes for wet-lab or animal work, and how much 2025 revenue rides on each use.

The finding: about 31.1% of 2025 computational biology revenue already depends on a regulator reading a model, and that share rises to about 33.4% by 2035 in the base case.

How was the computational biology seat model built, and which receipts check it?

The computational biology model multiplies 118,600 paid seats by USD 57,700 to reach USD 6.84 billion, then compounds 8.6% seat growth and 2.9% price growth to USD 20.8 billion in 2035. It covers five regions and rests on 14 sourced inputs.

The receipt arithmetic: 118,600 times USD 57,700 equals USD 6.84 billion. The cross-checks: the four disclosed vendor revenues sum to USD 828.6 million, or 12.1% of the market, consistent with a fragmented field. Certara’s software revenue per customer of USD 70,500 sits 22% above our blended seat price, as expected for a commercial-heavy base. Published 2025 estimates range from USD 7.18 billion to USD 8.1 billion; our USD 6.84 billion is 4.7% below the low end because computational genomics pipelines are excluded. Segment values sum to USD 6.84 billion in 2025 and within 0.1% of USD 20.8 billion in 2035.