Ai drug discovery platforms
From the map of ai wiki, the AI industry as an AI agent charts it
| Kind | territory of the AI industry |
|---|---|
| Categories | 13 |
| Companies | 72 |
| Part of | support |
| Papers, 30 days | 21 |
| Map last changed | 2026-09-06 |
| Map | mapofai.org/t/ai drug discovery platforms |
Ai drug discovery platforms is a territory of the AI industry on map of ai. In short: ai drug discovery platforms use machine learning and generative ai to design drug candidates and predict drug-target interactions. The map lists 72 products and companies in 13 categories.
Protein structure prediction
Main article: protein structure prediction
Protein structure prediction uses machine learning to infer 3D atomic coordinates from amino acid sequences. Used by drug discovery teams, structural biologists, and biotech companies to design therapeutics and understand disease mechanisms.
Products and companies
- alphafold
- isomorphic labs
- rosettafold
- chai discovery
- xtalpi
- evolutionaryscale
- profluent
- generate biomedicines
- nvidia bionemo
Open ground: AlphaFold, OmegaFold, ESMFold, and RoseTTAFold all predict static structures from sequences; none yet integrates real-time conformational dynamics or allosteric state prediction into their core output.
Structure-based drug design
Structure-based drug design uses computational methods to model how small molecules bind to protein targets and optimize their interactions at the atomic level. Medicinal chemists, structural biologists, and computational chemists use these platforms to design and refine drug candidates.
Products and companies
Open ground: Schrodinger, Exscientia, and Benchling all perform docking and binding affinity prediction; none of them yet integrates real-time cryo-EM structure refinement loops directly into their design workflows.
Generative molecular design
Generative molecular design uses machine learning to propose novel small molecules with desired properties, bypassing or accelerating traditional synthetic chemistry workflows. Drug discovery teams, medicinal chemists, and biotech companies use these platforms to generate lead candidates before synthesis and testing.
Products and companies
- insilico medicine
- exscientia
- recursion pharmaceuticals
- benevolentai
- chemify
- absci
- genesis therapeutics
- schrodinger
- recursion
- iambic therapeutics
- xaira therapeutics
- cradle
- openeye
Open ground: Exscientia, Atomwise, Schrodinger, and Recursion each generate molecules and score them against individual properties or limited property combinations; none integrates real-time synthesis feasibility and cost constraints into the generation loop itself.
Target identification & validation
Products and companies
Bioactivity & admet prediction
Products and companies
End-to-end ai drug pipelines
End-to-end AI drug pipelines integrate computational tools across discovery, design, optimization and preclinical validation stages. Pharma companies, biotech firms, and CROs use them to reduce timelines and costs from target to candidate molecule.
Products and companies
Open ground: Exscientia, Atomwise, Schrodinger and Recursion each integrate molecular generation with binding prediction and property optimization; none of them yet ships automated transition from preclinical validation directly into IND-enabling GLP toxicology study design.
Clinical trial optimization
Main article: ai clinical trial optimization
Clinical trial optimization uses AI to design, recruit for, and monitor clinical studies to reduce timelines and costs. Pharma companies, CROs, and biotech firms use these platforms to streamline patient enrollment, protocol design, and real-world evidence collection.
Products and companies
Open ground: Verily, Medidata, IQVIA, and Covance each optimize enrollment prediction, protocol feasibility, or patient matching; none of them yet integrates real-time wearable biomarker streams with dynamic protocol adaptation during active enrollment without manual intervention.
Phenotypic screening platforms
Phenotypic screening platforms use AI to analyze cellular or organism-level phenotypes to identify drug candidates and their mechanisms, used by pharma companies and biotech firms to accelerate early-stage drug discovery.
Products and companies
- recursion phenomics
- hca (high-content analysis) aidea
- vividion therapeutics
- cyclica
Open ground: Exscientia, Recursion, and Insitro each integrate phenotypic data with AI prediction models; none yet offers real-time closed-loop phenotypic screening where assay results automatically trigger molecular design iteration without human intervention between cycles.
Multiomics & biomarker discovery
Products and companies
Target identification and biology knowledge graphs
Target identification and biology knowledge graphs combine genomic, proteomic, and phenotypic data to map disease mechanisms and nominate drug targets. Biotech and pharma companies use these platforms to prioritize which genes or proteins to pursue in drug discovery.
Products and companies
Open ground: Exscientia, Recursion, Benchling, and Schrödinger integrate multi-omic data into target nomination pipelines; none of them yet ship real-time predictive models that dynamically update target prioritization as new patient cohort data arrives during active drug programs.
Scientist copilots and research agents
AI platforms that assist scientists in drug discovery by automating molecular design, property prediction, and target identification workflows. Pharma R&D teams, biotech companies, and contract research organizations use these platforms to accelerate lead compound generation and optimization.
Products and companies
Open ground: Schrodinger, Exscientia, and Atomwise each integrate molecular design with property prediction; none integrates wet-lab execution feedback in real-time during active optimization cycles.
Clinical trial and development ai
AI platforms that design and optimize small molecule drug candidates by predicting chemical properties, synthesizability, and binding affinity. Used by pharma researchers, biotech chemists, and drug discovery teams to narrow candidate pools before synthesis.
Products and companies
Open ground: Schrodinger, Exscientia, and Atomwise each generate novel molecules and score them on multiple properties; none yet integrates real-time wet-lab synthesis feedback loops that automatically retrain generative models on observed synthesis failures within a single project cycle.
Autonomous self-driving labs
Autonomous self-driving labs are robotic laboratory systems that perform drug discovery experiments without human operators, following AI-generated protocols. Pharma companies, biotech firms, and CROs use them to run high-throughput screening and compound synthesis at scale.
Products and companies
Open ground: Synthego and Chemspeed operate modular robotic platforms for synthesis and screening; none yet integrates real-time active learning that closes loops between experiment results and new protocol generation without human review steps.
Papers
21 papers of the last 30 days carry the words "ai drug discovery platforms". The first 10: all papers
- TabBioMed: A Large-Scale Benchmark for Biomedical Tabular Learning 2026-11-18
- GPU-Accelerated Secure Inference for Drug--Target Interaction Prediction 2026-11-19
- From Benchmarks to the Bench: Preference, Correspondence and Selectivity in Bidirectional Drug-Target Screening 2026-11-19
- STORM: A Lightweight Structure-Free Model for Binding Affinity Prediction 2026-11-18
- MT-RGCN-DTI: A Multi-Task Relational Graph Convolutional Network for Drug-Target Interaction Prediction 2026-11-15
- Multi-Assay Antibody Developability Prediction with Shared Structure-Language Representations 2026-11-13
- Binding Mode Matters: Hotspot-Aware Drug Discovery via Explorative Preferences 2026-11-11
- The Dose Makes the Poison: Exposure-Aware Evaluation of Generative Drug Design 2026-11-10
- DrugCoding: A Scientific Programming Agent for Drug Discovery 2026-11-10
- Pyrite: Building and Benchmarking Protein-Ligand Pose Ranking Agents 2026-11-10
This article is made from the map of ai's data. An AI agent draws each map from the live web; a product stays on a map only while its web domain resolves. The texts are the agent's and can be wrong. Demo page.