AI Rank
← 开源
ai4s-research

awesome-ai-for-science

A curated list of awesome AI tools, libraries, papers, datasets, and frameworks that accelerate scientific discovery — from physics and chemistry to biology, materials, and beyond.

ListsDomain-specific
在 GitHub 打开
增长势头
+624 小时新增 Star+0.3%
2.02k
Star
248
Fork
+14
本周
2
贡献者
创建于 2025-10-04 · 更新于 2026-10-07 · 今日第 1576 名
主要开发者
ai-boostChaoYue0307
README

✨ Awesome AI for Science (AI4Science) ✨

Awesome AI for Science Banner

A curated list of awesome AI tools, libraries, papers, datasets, and frameworks that accelerate scientific discovery across all disciplines.






[Deutsch](https://zdoc.app/de/ai-boost/awesome-ai-for-science) | 
[English](https://zdoc.app/en/ai-boost/awesome-ai-for-science) | 
[Español](https://zdoc.app/es/ai-boost/awesome-ai-for-science) | 
[français](https://zdoc.app/fr/ai-boost/awesome-ai-for-science) | 
[日本語](https://zdoc.app/ja/ai-boost/awesome-ai-for-science) | 
[한국어](https://zdoc.app/ko/ai-boost/awesome-ai-for-science) | 
[Português](https://zdoc.app/pt/ai-boost/awesome-ai-for-science) | 
[Русский](https://zdoc.app/ru/ai-boost/awesome-ai-for-science) | 
[中文](https://zdoc.app/zh/ai-boost/awesome-ai-for-science)





[![Awesome](https://awesome.re/badge.svg)](https://awesome.re)
[![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](https://opensource.org/license/MIT)
[![GitHub stars](https://img.shields.io/github/stars/ai-boost/awesome-ai-for-science.svg?style=social&label=Star)](https://github.com/ai-boost/awesome-ai-for-science)
[![GitHub forks](https://img.shields.io/github/forks/ai-boost/awesome-ai-for-science.svg?style=social&label=Fork)](https://github.com/ai-boost/awesome-ai-for-science)

AI is revolutionizing scientific research - from drug discovery and materials design to climate modeling and astrophysics. This repository collects the best resources to help researchers leverage AI in their work.

📚 Contents

🧪 AI Tools for Research

Literature & Knowledge Management

  • Semantic Scholar - AI-powered academic search (Allen AI)
  • arXiv - Open-access repository of electronic preprints and postprints
  • OpenAlex - Open catalog of scholarly papers and authors
  • CORE - Aggregator of open access research papers
  • Connected Papers - AI-powered visual graph for exploring academic papers and discovering connected research through citation networks and semantic similarity
  • PaSa (ByteDance) - Advanced paper search agent powered by large language models, autonomously invoking search tools, reading papers, and selecting references to deliver comprehensive and accurate results for complex scholarly queries (1.5K+ stars, Apache 2.0, 2024)
  • paper-search-mcp - MCP server, CLI, and agent skills for searching and downloading academic papers from multiple open sources (arXiv, PubMed, bioRxiv, Semantic Scholar, OpenAlex, CORE, Europe PMC, etc.) with unified, deduplicated, LLM-friendly retrieval and an OA-first download fallback chain (OpenAGS, 1.9K+ stars, MIT License, 2025)

Data Analysis & Visualization

  • PandasAI - Conversational data analysis using natural language
  • DeepAnalyze - First agentic LLM for autonomous data science with end-to-end pipeline from data to analyst-grade reports
  • AutoViz - Automated data visualization with minimal code
  • Chat2Plot - Secure text-to-visualization through standardized chart specifications

Data Labeling & Annotation

  • Label Studio - Multi-type data labeling and annotation tool
  • Snorkel - Programmatic data labeling and weak supervision

Research Workbench & Plugins

  • Claude Scientific Skills - Comprehensive collection of 125+ ready-to-use scientific skill modules for Claude AI across bioinformatics, cheminformatics, clinical research, ML, and materials science
  • GDM Science Skills - Google DeepMind's official collection of agentic science skills accelerating scientific workflows with better grounding and higher token efficiency, integrating insights from AlphaGenome, AFDB, UniProt and 30+ other databases and tools (2026)
  • Scientific Agent Skills - Turn any AI agent into an AI Scientist. The #1 Agent Skills library for science with 140+ ready-to-use skills and 100+ scientific databases covering biology, chemistry, medicine, and drug discovery. Compatible with Cursor, Claude Code, Codex, Antigravity, and the open Agent Skills standard (K-Dense-AI, 26K+ stars, 2025)
  • SciAgent-Skills - 197 bioinformatics and life science skills for Claude Code and AI agents, achieving 92.0% accuracy on BixBench. Covers RNA-seq, single-cell analysis, drug discovery, proteomics, and more. Powers OmicsHorizon (195+ stars, 2026)
  • Medical Research Skills - Curated library of 550+ medical research agent skills spanning evidence insights, protocol design, omics/clinical data analysis, and academic writing; each skill is reviewed through MedSkillAudit and compatible with Claude Code, Codex, Open Code, OpenClaw, and SKILL.md-compatible agents (AIPOCH, 1.2K+ stars, MIT License, 2026)
  • bioSkills - Collection of SKILLS.md guiding AI coding agents (Claude Code, OpenAI Codex, Google Gemini, OpenCode, OpenClaw) through common bioinformatics workflows from basic sequence manipulation to advanced analyses such as single-cell RNA-seq and population genetics; evaluated on the Bio-Task Bench dataset (GPTomics, 969+ stars, MIT License, 2026)

📄 Paper→Poster / Slides / Graphical Abstract

Poster Generation

  • Paper2Poster - Multi-agent system with Parser-Planner-Painter architecture converting paper.pdf to editable poster.pptx, outperforms GPT-4o with 87% fewer tokens
  • mPLUG-PaperOwl - Multimodal LLM for scientific charts and diagrams understanding/generation

Slides & Presentation Generation

  • Auto-Slides - Multi-agent academic paper to high-quality presentation slides with interactive refinement
  • PPTAgent - Beyond text-to-slides generation with PPTEval multi-dimensional evaluation (EMNLP 2025)
  • paper2slides - Transform arXiv papers into Beamer slides using LLMs
  • PaperToSlides - AI-powered tool that automatically converts academic papers (PDF) into presentation slides
  • pdf2slides - Convert PDF files into editable slides with three lines of code
  • SlideDeck AI - Co-create PowerPoint presentations with Generative AI from documents or topics
  • AI Multi-Agent Presentation Builder - Azure Semantic Kernel multi-agent PPT generation reference

Video & Media Generation

  • Paper2Video - First benchmark for automatic video generation from scientific papers (NeurIPS 2025)
  • paper2video - Transform arXiv research papers into engaging presentations and YouTube-ready videos

Website & Interactive Content Generation

  • Paper2All - AI-powered pipeline converting papers into interactive websites, posters, and multimedia presentations with "Let's Make Your Paper Alive!" philosophy

Figure & Illustration Generation

  • PaperBanana - Automated academic illustration generation for AI scientists, converting research papers into publication-ready figures using VLMs and diffusion models with iterative refinement (PKU & Google Research, 6.2K+ stars, 2026)

Chart & Visualization Generation

Note: For comprehensive chart understanding and code generation tools, see 📊 Chart Understanding & Generation section

📊 Chart Understanding & Generation

Chart-to-Code & Reproducibility

Scientific Visualization Tools

  • Chat2Plot - Secure text-to-visualization through standardized chart specifications
  • AutoViz - Automated data visualization with minimal code
  • PlotlyAI - AI-powered data visualization and dashboard creation
  • Flint (Microsoft) - Visualization intermediate language that lets AI agents create expressive, polished charts from simple, human-editable specs, compiling the same input to 30+ chart types across Vega-Lite, ECharts, and Chart.js with an MCP server for agent integration (1.9K+ stars, MIT License, 2026)

🔄 Paper-to-Code & Reproducibility

Automated Code Generation

  • Paper2Code - Automated code generation from machine learning research papers into runnable implementations (4.5K+ stars, 2025)
  • Paper2Agent - Multi-agent system automatically transforming research papers into interactive AI agents with MCP server generation, tutorial auto-detection, and benchmark extraction (2.2K+ stars, MIT License, 2025)
  • AutoP2C - LLM agent framework generating runnable repositories from academic papers
  • ResearchCodeAgent - Multi-agent system for automated codification of research methodologies
  • ToolMaker - Convert papers with code into callable agent tools

Experiment Automation

  • BioProBench - Comprehensive benchmark for automatic evaluation of LLMs on biological protocols and procedural understanding
  • Alhazen - Extract experimental metadata and protocol information from scientific documents

📋 Scientific Documentation & Parsing

High-Performance Document Processing

  • MinerU (2024/2025) - SOTA multimodal document parsing with 1.2B parameters outperforming GPT-4o, converts PDFs to LLM-ready Markdown/JSON
  • MinerU-Diffusion (OpenDataLab, ECCV 2026) - Diffusion-based document OCR framework replacing autoregressive decoding with block-level parallel diffusion decoding, enabling high-accuracy text recognition in scientific PDFs (613+ stars, MIT License)
  • OpenDataLoader PDF (OpenDataLoader, 2025) - Open-source PDF parser for AI-ready data, converting PDFs into Markdown/JSON/HTML/Tagged PDF with layout analysis and reading-order detection; ranks #1 overall on extraction benchmarks with deterministic bounding boxes and hybrid AI mode (26K+ stars, Apache 2.0)
  • PDF-Extract-Kit (2024) - Comprehensive toolkit for high-quality PDF content extraction with layout detection, formula recognition, and OCR
  • Docling (IBM, AAAI 2025) - Multi-format (PDF/DOCX/PPTX/HTML/Images) → structured data (Markdown/JSON) with layout reconstruction, table/formula recovery
  • Nougat (Meta AI) - Neural optical understanding for academic documents, transforms scientific PDFs to Markdown with mathematical formula support
  • olmOCR (AllenAI) - Toolkit for linearizing academic PDFs into LLM-ready text with high accuracy and structure preservation, optimized for scientific literature extraction
  • PaddleOCR 3.0 (2024/2025) - Advanced OCR with PP-StructureV3 document parsing, 13% accuracy improvement, supports 80+ languages
  • Unstructured - Production-grade ETL for transforming complex documents into structured formats, with open-source API
  • Marker - High-accuracy PDF→Markdown/JSON/HTML conversion, specialized for tables/formulas/code blocks with benchmark scripts
  • S2ORC doc2json (AllenAI) - Large-scale PDF/LaTeX/JATS parsing to standardized JSON for millions of papers
  • GROBID - Machine learning software for extracting structured metadata from scholarly documents
  • Science-Parse / SPv2 (AllenAI) - Parse scientific papers to structured fields (title/author/sections/references)

Production Pipelines & Data Preparation

Figure & Table Extraction

  • PDFFigures2 - Extract figures, tables, captions, and section titles from scholarly PDFs
  • TableBank - Large-scale table detection and recognition dataset with pre-trained models

Scientific Text Processing & NLP

  • scispacy (AllenAI) - Full spaCy pipeline and models for scientific/biomedical documents, enabling named entity recognition, abbreviation resolution, and UMLS linking for scientific literature mining (1.9K+ stars, Apache 2.0)

Scientific Literature RAG & Analysis

  • PaperQA2 - High-accuracy RAG for scientific PDFs with citation support, agentic RAG, and contradiction detection
  • OpenScholar - Retrieval-augmented LM synthesizing scientific literature from 45M papers with human-expert-level citation accuracy, outperforming GPT-4o by 5% on ScholarQABench (Nature 2026, UW & Ai2)
  • Valsci - Self-hostable scientific claim-verification and literature-review tool combining Semantic Scholar retrieval, bibliometric scoring, and LLM-based evidence synthesis for large-batch validation workflows
  • paper-reviewer - Generate comprehensive reviews from arXiv papers and convert to blog posts
  • STORM - LLM agent system synthesizing Wikipedia-like long-form research articles from scratch through multi-perspective question asking, web retrieval, and citation-grounded report generation, with Co-STORM extension for collaborative human-LLM knowledge curation conversations (Stanford OVAL, NAACL 2024 & EMNLP 2024)

🧰 Research Workbench & Plugins

Interactive Research Environments

  • Jupyter AI (JupyterLab Extension) - Official Jupyter extension with %%ai magic commands and sidebar chat assistant, connecting multiple model providers and local inference
  • Notebook Intelligence (NBI) - AI coding assistant for JupyterLab with agent mode, supporting arbitrary LLM providers (2025+)
  • Google Colab AI Features - Integrated AI assistance for data science and research notebooks
  • OpenAI4S - Open-source hybrid scientific research agent and workbench replicating Claude Science, combining JSON tool orchestration with persistent Python/R Code-as-Action kernels, 604 bundled science skills, MCP connectors, sandboxed local execution, and multi-provider LLM support for end-to-end scientific workflows (PKU–YuanKong Intelligence, 377+ stars, MIT License, 2026)
  • OpenBioMed - Open-source biomedical AI platform integrating multimodal foundation models (BioMedGPT, PharmolixFM, LangCell) with agentic workflows and 45+ Claude Code skills for drug discovery, protein engineering, and single-cell omics analysis (PharMolix & Tsinghua AIR, 1K+ stars, 2023-2026)
  • AutoR - Human-centered research OS with terminal-first harness and local browser Studio, turning research work into reproducible artifact-backed runs through a 9-stage workflow with human approval gates, resume/rollback controls, and venue-aware manuscript packaging (1K+ stars, 2026)
  • ScholarAIO - Agent-agnostic research infrastructure providing AI agents with a structured scientific workspace for deep PDF parsing, hybrid semantic/keyword literature search, citation-graph analysis, topic discovery, and academic writing workflows; natively integrates with Claude Code, Codex, Cursor, Cline, and AgentSkills.io (530+ stars, MIT License, 2026)
  • BioMCP - Biomedical Model Context Protocol (MCP) server unifying literature search across PubMed/Europe PMC, entity pivoting across genes/variants/drugs/diseases/pathways/proteins, local study analytics, and Claude Code/Codex integration for agentic biomedical research (531+ stars, MIT License, 2025-2026)
  • MATLAB Agentic Toolkit - Official MathWorks toolkit connecting AI agents to MATLAB via the MATLAB MCP Server and curated skills, enabling trusted engineering and scientific computing workflows with idiomatic code generation, testing, and error diagnosis in Claude Code, GitHub Copilot, OpenAI Codex, and Gemini CLI (686+ stars, BSD-3-Clause, 2026)
  • BioNeMo Agent Toolkit (NVIDIA) - Turn any AI agent into a life science expert with NVIDIA BioNeMo skills, enabling agentic workflows for drug discovery, protein engineering, and biomolecular design (329+ stars, Apache 2.0 / CC-BY-4.0, 2026)
  • open-science - Local-first, open-source AI workbench for scientists — an open alternative to Claude Science (by ai4s-research, maintainers of this list; TypeScript, MIT, 2026)
  • OpenScience (Synthetic Sciences) - Open-source AI workbench for scientific research that automates the full research loop — literature review, hypothesis generation, code writing, experiment execution, database querying, and report writing — with 290+ skills, specialized research agents, and a browser-based workspace (1453+ stars, Apache 2.0, 2026)
  • Claude Scholar - Semi-automated research assistant for academic research and software development, supporting Claude Code, Codex CLI, Kimi Code CLI, and OpenCode across ideation, coding, experiments, writing, and publication (Galaxy-Dawn, 4.5K+ stars, MIT License, 2026)
  • K-Dense BYOK - Free, open-source desktop AI research assistant that runs locally and turns natural-language requests into real data analysis, literature search, figure generation, and manuscript review; ships with 149 scientific skills, 326 workflow templates, and 229 databases across genomics, proteomics, drug discovery, and materials science, plus a living lab notebook, 60+ scientific file previews, and LaTeX editing (K-Dense-AI, 908+ stars, MIT License, 2026)
  • Science Superpowers (K-Dense-AI) - Composable computational-science methodology skills for AI research agents emphasizing pre-registration, reproducible workspaces, and red-team review to guard against p-hacking and HARKing; zero third-party dependencies and runs with any agent harness plus a POSIX shell (281+ stars, MIT License, 2026)
  • Wisp Science - Open-source, local-first desktop AI research workbench for scientific computing with Python/R, MCP bioinformatics tools, SSH/WSL/GPU runtimes, and OpenAI/Anthropic models (857+ stars, 2026)
  • Academic Research Skills (ARS) - Comprehensive Claude Code skill suite covering the full academic pipeline from deep research and paper writing to multi-perspective peer review, revision, and finalization; features multi-agent teams, PRISMA systematic review, style calibration, claim-level citation audits, integrity gates, and human-in-the-loop safeguards (38K+ stars, CC BY-NC 4.0, 2026)
  • Qinyan Academic Skills - Curated, multilingual library of 182 installable AI agent skills for end-to-end academic research spanning literature discovery, scientific writing, grant development, bioinformatics, drug discovery, clinical research, machine learning, and data analysis (779+ stars, MIT License, 2026)
  • SkillOpt (Microsoft, 2026) - Text-space optimizer that treats agent skill documents as trainable parameters for frozen LLMs, using scored rollouts and held-out validation gates to iteratively improve reusable natural-language skills; includes SkillOpt-Sleep for nightly self-evolution and improves accuracy across Claude Code, Codex, Copilot, and direct-chat harnesses, making it a meta-tool for evolving scientific agent skill workflows (15.5K+ stars, MIT License, PyPI)
  • Open Science (AIPOCH) - Open-source, local-first, model-agnostic AI research workbench for reproducible scientific discovery; runs Python/R notebooks, searches the web, calls scientific data connectors, and produces inspectable reports, tables, and figures in a self-hosted desktop workspace (1.5K+ stars, Apache 2.0, 2026)
  • OmicsClaw - Local-first, conversational AI research partner for multi-omics analysis with CLI, desktop app, and 95+ reproducible skills; keeps raw data local while routing natural-language requests to Python/R/CLI tools with persistent memory, autonomous analysis paths, and multi-method consensus workflows (TianGzlab, 155+ stars, Apache 2.0, 2026)
  • MedgeClaw - Open-source AI research assistant for biomedicine — chat to run RNA-seq, drug discovery, clinical analysis, and more; built on OpenClaw and Claude Code with 140 K-Dense scientific skills, real-time dashboard, and RStudio/JupyterLab integration (xjtulyc, 669+ stars, 2026)

Literature Management Plugins

  • llm-for-zotero - Research agent system deeply integrated with Zotero supporting Agent Mode, skills, multi-model backends (OpenAI-compatible, Claude Code, WebChat, Codex), and MinerU PDF parsing for literature Q&A, summarization, figure inspection, and source comparison (1.3K+ stars, 2026)
  • PapersGPT for Zotero - Multi-PDF conversation, retrieval, and citation in Zotero with commercial/local models (Ollama), MCP support
  • Zotero-GPT (MuiseDestiny) - Classic open-source plugin for document Q&A and summarization within Zotero
  • Better BibTeX for Zotero - Enhanced citation key management and LaTeX integration

Scientific Writing & Collaboration

  • Notion AI - AI-powered research note-taking and knowledge management
  • Obsidian Smart Connections - AI-powered note linking and research graph navigation
  • Research Rabbit - AI-powered literature discovery and research network mapping
  • SciWrite - Agent skill for AI-assisted scientific manuscript writing review distilled from Stanford's Writing in the Sciences course, performing five sequential editorial audit passes on clarity, voice, structure, consistency, and integrity (2026)
  • PaperSpine - Motivation-driven academic writing system for Claude Code, Codex, OpenClaw, and Hermes CLI that learns from strong papers, builds evidence-aware central-argument blueprints, and rewrites manuscripts with revision matrices and LaTeX-safe audits (4.9K+ stars, MIT License, 2026)
  • Claude Prism - Offline-first scientific writing workspace powered by Claude, integrating LaTeX, Python, and 100+ scientific skills with local execution, Zotero integration, and privacy-focused design (2026)

🕸 Knowledge Extraction & Scholarly KGs

Knowledge Graph Construction

  • iText2KG - Incremental knowledge graph construction using LLMs with entity extraction and Neo4j visualization
  • GraphGen - Knowledge graph-guided synthetic data generation for LLM fine-tuning, achieving strong performance on scientific QA (GPQA-Diamond) and math reasoning (AIME)
  • KoPA - Structure-aware prefix adaptation for integrating LLMs with knowledge graphs (ACM MM 2024)
  • Scholarly KGQA - LLM-powered question answering over scholarly knowledge graphs (ArXiv paper)
  • SciAtlas - Large-scale knowledge graph and pip-installable client for literature-grounded automated scientific research, connecting papers, authors, institutions, venues, keywords, citations, and a four-level research taxonomy across medicine, social sciences, engineering, computer science, materials science, and more (ZJU NLP, arXiv 2026, 136+ stars, MIT License)

Knowledge Graph Resources

  • Awesome-LLM-KG - Comprehensive collection of papers on unifying LLMs and knowledge graphs

🤖 Research Agents & Autonomous Workflows

Autonomous Research Systems (2023-2025 Breakthroughs)

  • FunSearch (DeepMind, Nature 2023) - First system to make novel, verifiable scientific discoveries by pairing LLMs with evolutionary search, solving open problems in combinatorics (cap set problem) and discovering faster matrix multiplication algorithms
  • OpenEvolve - Open-source implementation of AlphaEvolve's evolutionary coding agent paradigm, enabling LLMs to autonomously discover and optimize algorithms through iterative evolution, matching the approach behind DeepMind's breakthrough matrix multiplication discovery (6.2K+ stars, 2025)
  • Darwin Gödel Machine (Sakana AI) - Open-ended self-improving agent that iteratively rewrites its own codebase and empirically validates each mutation on coding benchmarks (SWE-bench, Polyglot), demonstrating open-ended evolution where agents improve their ability to improve themselves, diverging into a population of diverse specialists (arXiv 2505.22954, 2.3K+ stars, Apache 2.0, 2025)
  • SkyDiscover - Modular framework for AI-driven scientific and algorithmic discovery, providing a unified interface for implementing, running, and fairly comparing discovery algorithms across 200+ optimization tasks; introduces AdaEvolve and EvoX adaptive/evolutionary algorithms and natively supports OpenEvolve, GEPA, and Harbor-format benchmarks (skydiscover-ai, 568+ stars, Apache 2.0, 2026)
  • EvoMaster (SJTU SAI, arXiv 2026) - Foundational auto-research agent framework for agentic science at scale, providing modular agent construction, run-level self-evolution, and multiple SciMaster domain agents (ML-Master, X-Master, Browse-Master); outperforms general-purpose agents across authoritative benchmarks including the OpenAI Frontier Science Benchmark (206+ stars, Apache 2.0, 2026)
  • Virtual Lab (Stanford Zou Group, Nature 2025) - AI-human collaborative research platform where a human researcher works with a team of LLM agents via team and individual meetings to perform scientific research; demonstrated by designing new SARS-CoV-2 nanobodies with wet-lab validation
  • AI Co-Scientist (Google DeepMind, Nature Medicine 2026) - Multi-agent AI research partner that generates, reviews, ranks, and evolves research hypotheses alongside scientists, with experimental validation in biomedicine and other domains (2026)
  • Hyra (Tencent Hunyuan, 2026) - Hunyuan Research Agent for autonomous open-ended discovery across AI4Science, mathematics, and engineering, releasing reproducible solution artifacts for autocorrelation constants, Erdős problems, PARP1 docking, qubit routing, and record-breaking packing problems (Hyra-results, 112+ stars, Apache 2.0)
  • The AI Scientist (SakanaAI) - First fully autonomous open-ended scientific discovery system with official implementation: hypothesis→experiment→writing→review simulation (13.8K+ stars, 2024)
  • The AI Scientist v2 (SakanaAI) - Official implementation of the second-generation fully autonomous scientific discovery system, extending the original with agentic tree search and reduced template dependency to achieve workshop-level accepted papers (6.7K+ stars, 2025)
  • The AI Scientist v1 (2024) - First fully autonomous research system: hypothesis→experiment→writing→review simulation
  • The AI Scientist v2 (2025) - Enhanced with Agentic Tree Search, reduced template dependency, first workshop-level accepted paper
  • FAROS (OpenNSWM-Lab) - Foundation AutoResearch Operating System: blueprint-driven runtime for orchestrating AI research workflows from idea generation and experiments to paper writing and peer review (OpenNSWM-Lab, 2.4K+ stars, 2026)
  • DeepScientist - First system progressively surpassing human SOTA on frontier AI tasks (183.7%, 1.9%, 7.9% improvements), month-long autonomous discovery with 20,000+ GPU hours
  • ASI-Arch (GAIR-NLP, arXiv 2025) - Autonomous multi-agent research loop for model architecture discovery that ran 1,773 experiments over 20,000 GPU hours and produced 106 state-of-the-art linear-attention architectures, surpassing human-designed baselines including Mamba2 and DeltaNet (1.1K+ stars, Apache 2.0)
  • Kosmos - Extended autonomy AI scientist with 200 parallel agent rollouts, 42K lines of code execution, 1.5K papers analyzed per run, achieving 79.4% accuracy and 7 scientific discoveries (Edison Scientific)
  • AlphaResearch - Autonomous algorithm discovery combining evolutionary search with peer-review reward models, achieving best-known performance on circle packing problems
  • AutoResearchClaw - Fully autonomous research from idea to paper with multi-agent debate, citation verification, and OpenClaw integration (11K+ stars, 2026)
  • ARIS (Auto-Research-In-Sleep) - Lightweight Markdown-only skills for autonomous ML research with cross-model review loops, idea discovery, and experiment automation; no framework lock-in, works with Claude Code, Codex, OpenClaw, or any LLM agent (12.8K+ stars, MIT License, 2026)
  • Arbor - Generalist autonomous research agent that grows a hypothesis tree to optimize any measurable task, beating Claude Code and Codex by 2.5× on the same compute budget across BrowseComp, Terminal-Bench 2.0, math reasoning, and MLE-Bench Lite; supports native CLI, keyless Claude Code/Codex integration, and an MCP tool server (RUC-NLPIR, 866+ stars, Apache 2.0, 2026)
  • NanoResearch - End-to-end autonomous AI research engine that turns an idea into a complete LaTeX paper by dispatching real computational experiments to local GPUs or SLURM clusters, collecting actual results, generating figures/tables, and writing a data-grounded manuscript rather than LLM hallucinations (OpenRaiser, 1.5K+ stars, MIT License, 2026)
  • ScienceClaw - Self-evolving AI research colleague built on OpenClaw with 285+ runtime-adaptive skills across 28+ disciplines, persistent cross-session research memory, and zero-hallucination citation protocols; agent autonomously writes new SKILL.md files based on research patterns without redeployment (828+ stars, MIT License, 2026)
  • ai4s-skills - Agent skills (SKILL.md + deterministic tools) for the AI4S workflow — topic exploration, literature survey, runnable experiments, publication-grade papers, and integrity audit, with every citation and number traceable to its source (by ai4s-research, maintainers of this list; MIT, 2026)
  • Denario (AstroPilot-AI, Agents4Science 2025) - Modular multi-agent scientific research assistant that automates idea generation, literature review, methodology design, code execution in Docker, visualization, LaTeX paper writing, and peer-review simulation across 10+ disciplines; winner of the NeurIPS 2025 Fair Universe Competition (573+ stars, GPL-3.0, 2025-2026)
  • AI-Researcher - Autonomous pipeline from literature review→hypothesis→algorithm implementation→publication-level writing with Scientist-Bench evaluation
  • Agent Laboratory - Multi-agent workflows for complete research cycles with AgentRxiv for cumulative discovery
  • AIDE (WecoAI, arXiv 2025) - LLM-driven machine learning engineering agent using agentic tree search to autonomously draft, debug and benchmark ML code; wins 4× more medals than the best linear agent on OpenAI's MLE-Bench (75 Kaggle competitions) (1.3K+ stars, MIT License)
  • RD-Agent (Microsoft) - Open-source LLM-powered R&D agent framework automating data-driven AI solution building through automated research, development, and evolution; achieves top open-source performance on MLE-Bench with dual Researcher-Developer agents and supports research copilot, data mining, Kaggle, and quant R&D workflows (13.6K+ stars, MIT License, 2025-2026)
  • CodeScientist (AllenAI) - End-to-end semi-automated scientific discovery system that designs, iterates, and analyzes code-based experiments via LLM-as-a-mutator over scientific articles and code examples; auto-creates, runs, and debugs experiment code in containers and writes meta-analysis reports (339+ stars, Apache 2.0)
  • InternAgent - Closed-loop multi-agent system from hypothesis to verification across 12 scientific tasks, #1 on MLE-Bench (36.44%)
  • freephdlabor - First fully customizable open-source multiagent framework automating complete research lifecycle from idea conception to LaTeX papers with dynamic workflows
  • AutoScientists (Harvard MIMS, arXiv 2026) - Decentralized self-organizing teams of AI agents for long-running computational scientific experimentation; agents critique each other's proposals before spending compute and share successes/failures to avoid redundant exploration, achieving +8.33% on BioML-Bench, 1.9× faster nanoGPT optimization, and +12.5% on ProteinGym ACE2-Spike (425+ stars, 2026)
  • ToolUniverse - Democratizing AI scientists by transforming any LLM into research systems with 600+ scientific tools (Harvard MIMS)
  • LabClaw - Skill operating layer for biomedical AI agents with 211 production-ready SKILL.md files across 7 domains (biology, pharmacology, medicine, data science, literature search), enabling modular dry-lab reasoning and protocol composition for Stanford LabOS-compatible agents
  • Robin - FutureHouse's end-to-end scientific discovery multi-agent system orchestrating literature search (Crow/Falcon) and data analysis (Finch) agents, first AI-generated drug discovery identifying ripasudil as novel dry AMD therapeutic (2025)
  • Aviary - Language agent gymnasium for challenging scientific tasks including DNA manipulation, literature search, and protein engineering
  • Curie - Automated and rigorous experiments using AI agents for scientific discovery
  • POPPER - Automated hypothesis testing with agentic sequential falsifications
  • autoresearch - Andrej Karpathy's autonomous LLM research framework: AI agent runs overnight experiments on a real training setup, auto-editing code→5min training→evaluation in a loop, ~100 experiments per night on a single GPU
  • UniScientist - Universal scientific research intelligence covering 50+ disciplines, repositioning LLMs as cross-disciplinary generators with human experts as verifiers; 30B model outperforms Claude Opus and GPT on 5 research benchmarks
  • EvoScientist - Self-evolving AI scientist with 6 specialized sub-agents (plan/research/code/debug/analyze/write) and persistent memory, #1 on DeepResearch Bench II and AstaBench, supporting multi-provider LLMs and multi-channel deployment (Apache 2.0, 2026)
  • PantheonOS (Stanford, 2025) - Evolvable and privacy-preserving multi-agent framework automating, scaling, and accelerating data sciences with a particular focus on end-to-end single-cell biology analyses; features agentic code evolution, multi-agent team orchestration, distributed architecture, and a community marketplace with 1,000+ curated agents and skills (428+ stars)
  • CORAL (arXiv 2026) - Robust, lightweight infrastructure for multi-agent autonomous self-evolution, built for autoresearch; agents run in isolated git worktrees, share knowledge through a common state directory, and are scored by a grader daemon; natively integrated with Claude Code, Codex, Cursor Agent, OpenCode, and Kiro (672+ stars, Apache 2.0)
  • Science-Star (USTC AI4Science, 2025) - Open-source platform for building, extending, and experimenting with scientific agents, providing modular agent construction tools and standardized evaluation pipelines for accelerating autonomous scientific discovery research (748+ stars, MIT License)
  • SR-Scientist (ICLR 2026) - Scientific equation discovery with agentic AI, elevating LLMs from equation proposers to autonomous scientists that write code, analyze data, implement equations, and optimize based on experimental feedback; outperforms baselines by 6-35% across four science disciplines with robustness to noise and out-of-domain generalization (GAIR-NLP / SJTU, 49+ stars, Apache 2.0)
  • ARA (Agent-Native Research Artifact) - Research ecosystem for rigorous and trustworthy AI scientists — a protocol and skill bundle that makes autonomous research verifiable, crystallized, and observable through structured, machine-executable research artifacts and five agent skills for research management, compilation, verification, visualization, and publication (ARA-Labs, 447+ stars, MIT License, 2026)
  • XScientist - Local-first autonomous research system implementing a Git-like research protocol for long-running scientific discovery; explores competing explanations, executes experiments inside an isolation boundary, self-criticizes results, and exports the entire path as typed Agent-Native Research Artifacts (ARA) with exploration DAGs, claim-to-evidence anchors, content hashes, and re-execution hooks (126+ stars, Apache 2.0, arXiv 2026)
  • Scholar Loop - Autonomous multi-agent AI scientist that mirrors a PhD workflow: literature review → grounded hypothesis → real ML experiments → self-critique → write-up; features a deterministic harness with frozen-metric scoring, edit allowlists, and a verified registry to make reward-hacking and hallucination impossible, plus 108 unit tests runnable without API keys or GPUs (461+ stars, MIT License, 2026)
  • ResearchStudio (Microsoft) - AI co-author covering the entire research lifecycle — from an under-specified research direction to a published paper; includes ResearchStudio-Idea for evidence-grounded research ideation and ResearchStudio-Reel for turning finished papers into posters, narrated videos, blogs, and interactive reels; runs as skills on Claude Code and Codex (1.2K+ stars, MIT License, 2026)
  • Principia - Principle-first scientific idea discovery framework that extracts reusable principles from public literature and private research materials, composes them into traceable Idea Cards with prior-art comparisons, and exports validation-ready research packs; emphasizes inspectable scientific objects, risk disclosure, and falsification paths (ICML 2026, 411+ stars, MIT License)
  • Imbue Catalyst - Semi-autonomous AI scientist for scientific theory discovery and verifiable goal solving, using adversarial review-refinement loops and evolution-inspired candidate populations; integrates with Claude Code, Gemini CLI, Antigravity, and Codex harnesses (Imbue, 31+ stars, AGPL-3.0, 2026)

Evaluation & Benchmarking

  • ScienceAgentBench (ICLR 2025) - 102 executable tasks from 44 peer-reviewed papers across 4 disciplines with containerized evaluation
  • AIRS-Bench (Meta, 2026) - Benchmark quantifying end-to-end autonomous AI research abilities of LLM agents across 20 tasks from SOTA machine learning papers spanning NLP, code, math, biochemical modelling, and time series forecasting, with normalized score metrics against human SOTA and HuggingFace dataset
  • PaperBench (OpenAI, 2025) - Benchmark evaluating AI agents' ability to replicate 20 ICML 2024 Spotlight/Oral papers from scratch, with 8,316 gradable tasks and author-co-developed rubrics
  • PaperGuru (AutoTrustAI, 2026) - Lifecycle-Aware Memory (LAM) primitive and benchmark for long-horizon research agents, achieving 65.95% mean reproduction on PaperBench and 94.66% on SurveyBench through Capital Chunk Memory (CCM) with versioned content, structural multi-hop relevance, and provenance-grounded composition; 10 peer-reviewed acceptances at FSE/ICML/TOSEM/AEI/ICoGB (1.3K+ stars)
  • MLE-Bench (OpenAI, 2024) - Benchmark evaluating AI agents on 75 curated Kaggle-style ML engineering competitions with reproducible Docker-based grading harness, human baselines, and end-to-end task lifecycle, used as a primary benchmark for autonomous ML research agents (e.g., InternAgent #1 at 36.44%)
  • ScienceBoard (ICLR 2026) - Evaluating multimodal autonomous agents in realistic scientific workflows across real scientific software environments (KAlgebra, Celestia, Grass GIS, Lean 4, etc.) with VM-based evaluation infrastructure and agent trajectories
  • BuildArena - First physics-aligned interactive benchmark for LLM agents in engineering construction, designing rockets/cars/bridges in physics simulator with 3D spatial geometry library
  • SciTrust (2024) - Trustworthiness evaluation framework for scientific LLMs (truthfulness, hallucination, sycophancy)
  • SciCode - Research coding benchmark curated by scientists with 338 subproblems across 16 subdomains (physics, math, materials, biology, chemistry), evaluating LLMs on realistic scientific programming tasks with gold-standard solutions (NeurIPS 2024)
  • SciBench - College-level scientific problem-solving evaluation across multiple domains
  • NewtonBench (ICLR 2026) - First benchmark evaluating LLMs' ability to rediscover scientific laws through interactive experimentation across 324 tasks in 12 physics domains, featuring memorization-resistant metaphysical shifts of canonical laws (HKUST)
  • ResearchClawBench (InternScience, arXiv 2026) - Benchmark evaluating AI agents for end-to-end automated research from re-discovery to new-discovery, with 40 real-science tasks across 10 disciplines, curated datasets from published papers, and expert-curated multimodal rubrics (170+ stars, MIT License)
  • Terminal-Bench Science (Harbor Framework, 2026) - Benchmark evaluating AI agents on complex real-world scientific workflows in terminal environments across life, physical, earth, and mathematical sciences; featured on model cards for Claude Opus 4.7, GPT-5.5, and Gemini 3.1 Pro (200+ stars, Apache 2.0)

Academic Review & Evaluation

  • AgentReview - LLM agents simulating academic peer review ecosystems
  • LLM-Peer-Review - Web application for LLM-assisted manuscript review and annotation

Domain-Specific Research Agents

  • Aletheia - Google DeepMind's autonomous mathematics research agent powered by Gemini Deep Think, autonomously solving 4 open problems from 700 Erdős conjectures and generating complete research papers without human intervention (February 2026)
  • AlphaProof Nexus (Google DeepMind, arXiv 2026) - LLM-driven formal proof search system that pairs large language models with Lean verification to solve open mathematics problems; autonomously resolved 9 of 353 Erdős problems and 44 of 492 OEIS conjectures, with proofs and natural-language prose released for combinatorics, optimization, graph theory, algebraic geometry, and quantum optics collaborations (282+ stars, Apache 2.0)
  • Ten Proofs (OpenAI, 2026) - Lean 4 formalizations of ten major advances in mathematics and theoretical computer science, including improved sphere-packing bounds, non-sofic groups, a counterexample to Connes's rigidity conjecture, and quantum parallel repetition; released with the OpenAI paper and reasoning walkthroughs (57+ stars, Apache 2.0)
  • AlphaGeometry - DeepMind's Olympiad-level geometry theorem prover combining neural language model with symbolic deduction engine, AlphaGeometry2 solves 84% of IMO geometry problems (42/50) at gold-medalist level (Nature 2024)
  • Goedel-Prover-V2 - Strongest open-source automated theorem prover in Lean 4, 8B model matches DeepSeek-Prover-V2-671B at 84.6% MiniF2F, 32B model achieves 90.4% with self-correction, using scaffolded data synthesis and verifier-guided proof refinement (Princeton, 2025)
  • DeepSeek-Prover-V2 - DeepSeek's open-source large language model for formal theorem proving in Lean 4, integrating informal and formal mathematical reasoning through recursive subgoal decomposition and reinforcement learning powered by DeepSeek-V3, with open weights and ProverBench evaluation (2025)
  • LeanDojo - Open-source toolkit and benchmark for learning-based theorem proving in Lean, providing programmatic Lean interaction, a 98K+ theorem dataset extracted from 217 Lean projects, and ReProver—the first retrieval-augmented LLM-based theorem prover for Lean—with reproducible training pipelines underpinning much subsequent Lean prover research (Caltech & NVIDIA, NeurIPS 2023 Outstanding Paper, Datasets & Benchmarks)
  • Lean Copilot - LLMs as copilots for theorem proving in Lean 4, exposing native tactics (suggest_tactics, search_proof, select_premises) that embed language model inference and premise retrieval directly inside the Lean proof environment, supporting local CTranslate2/CUDA inference as well as remote model APIs for interactive and automated proof search (Caltech & NVIDIA, NeurIPS 2024, 1.2K+ stars)
  • MathCode - Terminal AI coding assistant with a built-in math formalization engine that converts plain-language math problems into Lean 4 theorems and attempts formal proofs; bundles a local Lean toolchain and WebUI for interactive mathematical reasoning (math-ai-org, 582+ stars, 2026)
  • TorchLean (lean-dojo, 2026) - First unified Lean 4 framework for neural-network specification, execution, and verification; tensor shapes are part of the types, models are executable Lean programs, and the same definitions can be used by training code, graph transformations, certificate checkers, and proofs with CPU/CUDA backends (123+ stars, MIT License)
  • Get Physics Done (PSI) - First open-source agentic AI physicist turning research questions into structured workflows with rigorous verification and multi-step analytical work for long-horizon physics projects; integrates with Claude Code, Codex, Gemini CLI, and OpenCode (804+ stars, Apache 2.0, 2026)
  • Foam-Agent (NeurIPS 2025) - End-to-end composable multi-agent framework for automating OpenFOAM-based CFD simulations from natural language prompts, managing meshing, case setup, execution, error correction, and post-processing; achieves 100% success rate on 110 FoamBench tasks with Claude Opus 4.6 through Architect-Input Writer-Runner-Reviewer agent collaboration with RAG-enhanced generation and MCP tool integration (RPI CSML, 242+ stars, MIT License)
  • AI CFD Scientist (RPI CSML, arXiv 2026) - Open-ended AI scientist for computational fluid dynamics that spans literature-grounded ideation, OpenFOAM execution via Foam-Agent, vision-language physics verification of rendered flow fields, source-code modification for new physical models, and figure-grounded LaTeX manuscript writing within a single inspectable workflow (43+ stars, Python)
  • Zephyrus (ICLR 2026) - First agentic framework for weather science, pairing an LLM with ZephyrusWorld (a code-execution environment exposing WeatherBench 2 data, geolocation, forecasting, simulation, and climatology tools) and ZephyrusBench (2,230 Q&A pairs across 49 weather-science tasks); outperforms text-only baselines by up to 44.2 percentage points (UC San Diego Rose-STL-Lab, 99+ stars, MIT License, 2026)
  • BioDiscoveryAgent - AI agent for biological discovery and research automation
  • Biomni - General-purpose biomedical AI agent integrating LLM reasoning with retrieval-augmented planning and code-based execution to autonomously execute diverse biomedical research tasks and generate testable hypotheses (Stanford SNAP, bioRxiv 2025)
  • BioAgents - AI scientist framework for autonomous deep research in biological sciences, combining literature analysis agents with data scientist agents to enable iterative scientific discovery through user feedback integration; achieves state-of-the-art performance on BixBench benchmark (48.78% open-answer, 64.39% multiple-choice) outperforming Kepler and GPT-5 (bio-xyz, arXiv 2601.12542, 160+ stars, 2025-2026)
  • SRAgent - LLM agents for working with the SRA (Sequence Read Archive) and associated bioinformatics databases, enabling natural language querying of high-throughput sequencing data and metadata across genomic repositories (Arc Institute, 169+ stars, 2024-2026)
  • STAgent - Multimodal LLM-based AI agent enabling deep research in spatial transcriptomics, automating analysis and interpretation of spatial gene expression data (Harvard LiuLab, bioRxiv 2025)
  • Camyla - Fully autonomous medical image segmentation research system that generates complete manuscripts end-to-end from datasets with zero human intervention, beating strongest baselines on 24 of 31 datasets and achieving T1-T2 tier manuscript quality in double-blind evaluations (USTC & Shanghai AI Lab, 2026)
  • MOOSE - Large Language Models for automated open-domain scientific hypotheses discovery (ACL 2024, ICML Best Poster)
  • ChemCrow - LLM agents for chemistry research with tool integration
  • Coscientist - Autonomous chemical experiment planning and execution
  • SciAgents - Bioinspired multi-agent intelligent graph reasoning system that autonomously traverses ontological knowledge graphs to generate, critique, and refine novel research hypotheses, demonstrated on bio-inspired materials discovery with cross-disciplinary connection mining (MIT Lamm Group, 2024)
  • TxAgent - AI agent for therapeutic reasoning across a universe of tools, achieving 92.1% accuracy in drug reasoning and outperforming GPT-4o by 25.8% (Harvard MIMS, 2025)
  • ATHENA-R1 (Harvard MIMS) - Reinforcement-learning-trained AI agent for treatment reasoning over a universe of 212 biomedical tools, performing multi-step evidence gathering and spawning parallel reasoning branches to reach evidence-grounded clinical decisions (55+ stars, MIT License, 2026)
  • ClawBio - First bioinformatics-native AI agent skill library enabling local-first, reproducible genomic and population-genetics research workflows built on OpenClaw (871+ stars, MIT License, 2026)

🏷 Data Labeling & Curation

Weak Supervision & Auto-Labeling

  • Snorkel - Programmatic data labeling and weak supervision for scientific datasets
  • PandasAI - Conversational data analysis and visualization using natural language
  • Cleanlab - Standard data-centric AI package for data quality and machine learning, automatically detecting label errors, outliers, and dataset issues to improve scientific dataset reliability and model performance (11K+ stars, MIT License)

⚗ Scientific Machine Learning

Neural Differential Equations

Chemical Reaction Networks & Systems Biology

  • Catalyst.jl - Chemical reaction network and systems biology interface for scientific machine learning (SciML), enabling high-performance, GPU-parallelized simulation and analysis of complex biochemical systems with O(1) solvers (SciML, 518+ stars, Julia)

Physics-Informed Neural Networks

  • DeepXDE - Deep learning library for solving PDEs
  • Lang-PINN - LLM-driven multi-agent system that builds trainable PINNs from natural language task descriptions, achieving 3-5 orders of magnitude MSE reduction and 50%+ execution success improvement (ICLR 2026)
  • PINNs - Physics-informed neural networks
  • NVIDIA PhysicsNeMo - Open-source framework for building physics-ML models at scale (renamed from Modulus, 2025)
  • PINA - Physics-Informed Neural networks for Advanced modeling in PyTorch
  • NeuroMANCER (PNNL) - PyTorch-based differentiable programming framework for physics-informed system identification, parametric constrained optimization, and model predictive control, integrating neural operators, neural ODEs, KANs, SINDy, and differentiable predictive control with 30+ tutorials (1.3k+ stars, BSD License)
  • SciANN - Keras-based scientific neural networks
  • NeuralPDE.jl - Physics-informed neural networks in Julia

Neural Operators & Model Discovery

  • DeepONet - Learning nonlinear operators
  • PySINDy - Sparse identification of nonlinear dynamics
  • PySR - High-performance symbolic regression for discovering interpretable scientific equations from data, multi-population evolutionary search with Python/Julia backend, widely used in physics and astronomy (Cambridge, NeurIPS 2023)
  • LLM-SR - Scientific equation discovery and symbolic regression using LLMs, combining code generation with evolutionary search (ICLR 2025 Oral)
  • PSRN - Parallel symbolic regression network evaluating millions of expressions on GPU with automated subtree reuse, Nature Computational Science cover article (MIT, 2026)
  • pykan - Kolmogorov-Arnold Networks with learnable activation functions on edges instead of fixed node activations, achieving strong performance in function fitting, PDE solving, and scientific discovery with enhanced interpretability as an alternative to MLPs (MIT, 16.3K+ stars, 2024)
  • Fourier Neural Operator - Learning operators in Fourier space
  • Poseidon - Efficient foundation models for PDEs with pretrained transformer-based neural operators and downstream task fine-tuning pipelines, HuggingFace integration for models and datasets (ETH Zurich CAMLab, arXiv 2024)
  • GAOT (NeurIPS 2025) - Geometry Aware Operator Transformer serving as an efficient and accurate neural surrogate for PDEs on arbitrary domains, combining geometric priors with transformer architectures for scientific computing (ETH Zurich CAMLab, 92+ stars)
  • TensorMesh (ETH Zurich CAMLab, arXiv 2026) - Fast, differentiable, JIT-free finite element library for PyTorch enabling GPU-native PDE solving with native autograd, tensorized assembly, and sparse linear algebra; part of the TensorGalerkin framework (218+ stars, Apache 2.0)
  • PhiFlow - Differentiable PDE solving framework for machine learning with built-in fluid simulation, supporting PyTorch/JAX/TensorFlow backends and enabling neural network training within physical simulations (TUM, MIT License)
  • exponax - Efficient differentiable n-dimensional PDE solvers built on JAX and Equinox, shipping 46+ built-in equations with Fourier spectral methods, exponential time differencing, and full auto-differentiation for physics-based deep learning workflows (MIT, 200+ stars, 2024)

Simulation-Based Inference

  • sbi - Python package for simulation-based inference enabling likelihood-free Bayesian parameter estimation from scientific simulators, with flexible interfaces for neural posterior estimation, sequential methods, and MCMC/variational backends (Mackelab, 825+ stars)

📖 Papers & Reviews

Foundational Papers

📊 Comprehensive Surveys & Reviews (2024-2025)

AI for Scientific Research

Scientific Large Language Models

Scientific Machine Learning

Uncertainty Quantification

Automation & Self-Driving Laboratories

Policy & Strategic Perspectives

  • Artificial Intelligence for Science (CSIRO 2022) - Landmark report analyzing AI adoption across 98% of scientific fields over 60 years
  • AI for Science 2025 (Fudan University & Nature 2025) - Comprehensive report on AI's transformative impact across 7 scientific fields, 28 research directions, and 90+ challenges
  • AI in science evidence review (European Scientific Advice 2024) - Policy-focused evidence review on AI's impact in research

🚀 AI Scientist & Autonomous Research (2024-2025 Breakthroughs)

Recent Advances & Domain Applications

📈 Evaluation & Benchmarking

🔬 Domain-Specific Applications

🧬 Biology & Medicine

Protein & Drug Discovery

  • CryoDRGN - Neural network-based cryo-EM heterogeneous reconstruction, modeling continuous 3D structure distributions from single-particle images, with CryoDRGN-ET extending to in-cell cryo-electron tomography (MIT CSAIL, Nature Methods 2021/2024)
  • ModelAngelo - Automatic atomic model building program for cryo-EM maps using deep learning, enabling rapid de novo protein structure determination from electron density with high accuracy (3DEM/EMBL, 169+ stars)
  • AlphaFold - Protein structure prediction
  • AlphaFold3 - AlphaFold 3 inference pipeline for unified biomolecular structure prediction of proteins, nucleic acids, small molecules, ions, and post-translational modifications (Google DeepMind, Nature 2024)
  • AlphaFold Server - Free, easy-to-use web platform by Google DeepMind and Isomorphic Labs for running AlphaFold 3 predictions of biomolecular structures and interactions, enabling researchers without local infrastructure to model proteins, nucleic acids, small molecules, ions, and post-translational modifications through a searchable proteome interface (2024)
  • AlphaProteo - Deep learning system for de novo design of high-affinity protein binders, achieving strong binding across diverse target classes including challenging intracellular proteins with significantly higher success rates than traditional wet-lab screening methods (Google DeepMind, Nature 2024)
  • AlphaPulldown - Automated pipeline for proteome-scale protein-protein interaction screening with AlphaFold-Multimer and AlphaFold 3, supporting flexible inputs (UniProt IDs, FASTA, residue regions, multimers, AF3 JSON features) and integrated downstream analysis for hit prioritization (Kosinski Lab, EMBL, Nature Protocols 2024, 317+ stars, GPL-3.0)
  • RareFold - Structure prediction and design of proteins with noncanonical amino acids, enabling AI-powered modeling of synthetic biology constructs and expanded genetic code systems (133+ stars, 2025)
  • ColabFold (2025 Updates) - AlphaFold/ESMFold accessible implementation with AF3 JSON export, database updates
  • OpenFold - Trainable, memory-efficient PyTorch reproduction and retraining of AlphaFold2 providing new insights into its learning dynamics and out-of-distribution generalization; widely used as the open-source AlphaFold2 backbone underpinning many downstream protein structure prediction and design pipelines (Columbia AlQuraishi Lab & OpenFold Consortium, Nature Methods 2024)
  • OpenFold3 - Fully open-source (Apache 2.0) biomolecular structure prediction reproducing AlphaFold3, free for academic and commercial use (Columbia AlQuraishi Lab & OpenFold Consortium, 2025)
  • Protenix - Trainable PyTorch reproduction of AlphaFold 3
  • HelixFold3 - Baidu's open-source reproduction of AlphaFold3 in PaddlePaddle, providing pretrained weights and inference pipelines for unified biomolecular structure prediction across proteins, nucleic acids, ligands, ions, and post-translational modifications within the PaddleHelix biocomputing platform (Baidu, bioRxiv 2024)
  • RoseTTAFold-All-Atom - All-atom biomolecular structure prediction for protein-nucleic acid-small molecule-metal ion complexes, enabling accurate modeling of covalent modifications and assemblies beyond proteins (Baker Lab, Science 2024)
  • Chai-1 - Multi-modal foundation model for biomolecular structure prediction (proteins, small molecules, DNA, RNA, glycans) achieving SOTA across benchmarks, with optional MSA/template support (Chai Discovery, 2024)
  • Chai-2 (Chai Discovery, 2025) - Next-generation multi-modal foundation model for biomolecular structure prediction, ranking #1 at CASP16 and outperforming AlphaFold 3 on the CASP16 dataset with large gains on antibody-antigen and protein-protein complexes; open weights with free web interface (2K+ stars repo)
  • IntelliFold - Controllable foundation model for general and specialized biomolecular structure prediction across proteins, nucleic acids, and complexes, featuring a public web server for interactive prediction workflows (IntelliGen AI, 223+ stars, Apache 2.0, 2025)
  • SimpleFold (Apple, arXiv 2025) - Flow-matching protein folding model using only general-purpose transformer layers, scaled to 3B parameters and trained on 8.6M+ distilled structures; challenges the reliance on complex domain-specific architectures and supports PyTorch and MLX backends with model sizes from 100M to 3B parameters (985+ stars, MIT License)
  • NeuralPLexer - State-specific protein-ligand complex structure prediction with a multi-scale deep generative model, enabling conformational state-aware modeling of molecular interactions (329+ stars, 2024)
  • Boltz - First fully open-source model achieving AlphaFold3-level accuracy with 1000x faster binding affinity prediction (MIT)
  • Boltz-2 (MIT & Recursion, 2025) - Next-generation biomolecular foundation model jointly predicting protein-ligand complex structures and binding affinities in a single framework; achieves FEP-level accuracy with ~0.62 Pearson correlation on FEP+ benchmark in ~20 seconds, outperforming all methods at CASP16 affinity challenge and doubling average precision in MF-PCBA hit-discovery screens (MIT License)
  • BoltzGen - De novo protein binder design via generative model, achieving nanomolar binding for 66% of novel targets tested (MIT, 2025)
  • Proteina-Complexa - Flow-based generative model for atomistic protein binder design with test-time optimization, SOTA on binder benchmarks (ICLR 2026 Oral, NVIDIA)
  • PXDesign (ByteDance, 2025) - Fast, modular, and accurate de novo design of protein binders based on the Protenix foundation model, achieving 17-82% nanomolar hit rates across diverse targets with 2-6× improvement over prior methods like AlphaProteo and RFdiffusion (229+ stars, Apache 2.0)
  • ODesign (OTeam-AI4S, 2025) - All-atom generative world model for all-to-all biomolecular interaction design, enabling cross-modality generation of proteins, nucleic acids, small molecules, and cyclic peptides with fine-grained epitope-level control and 2-4 orders of magnitude faster design throughput than modality-specific baselines (316+ stars, Apache 2.0)
  • OpenDDE (Aureka Research, 2026) - Open-source, all-atom biomolecular foundation model that turns co-folding into a scalable engine for structure prediction, design, and optimization across proteins, nucleic acids, and small molecules in drug discovery; ranked first on PXMeter-AB, FoldBench-AB, and 2026ARK-AB antibody-antigen benchmarks (263+ stars, Apache 2.0)
  • La-Proteina (NVIDIA) - Partially latent flow matching model for the joint generation of a protein's amino acid sequence and full atomistic structure, including both backbone and side chains (2025)
  • Proteina (NVIDIA, ICLR 2025 Oral) - Large-scale flow-based protein backbone generator utilizing hierarchical fold class labels for conditioning with a tailored scalable transformer architecture, enabling controllable de novo protein design (264+ stars)
  • xfold - Democratizing AlphaFold3: PyTorch reimplementation to accelerate protein structure prediction research
  • MegaFold - Cross-platform system optimizations for accelerating AlphaFold3 training with 1.73x speedup and 1.23x memory reduction
  • Graphormer - General-purpose deep learning backbone for molecular modeling
  • DiffDock - Diffusion-based molecular docking achieving SOTA blind docking performance, treating ligand pose prediction as generative diffusion over SE(3), with DiffDock-L update for improved generalization (MIT CSAIL, ICLR 2023)
  • GNINA - Deep learning framework for molecular docking extending AutoDock Vina with convolutional neural network scoring functions, achieving superior virtual screening enrichment and pose prediction across diverse target classes; widely adopted in pharmaceutical structure-based drug design (J. Cheminformatics, 915+ stars, actively maintained)
  • DynamicBind (NeurIPS 2024) - Deep equivariant generative model predicting ligand-specific protein-ligand complex structures with dynamic receptor conformational flexibility, enabling accurate docking for flexible protein targets
  • PLACER - Graph neural network operating entirely at the atomic level for protein-ligand conformational ensemble prediction and docking, generating diverse solutions through rapid stochastic denoising to model conformational heterogeneity (Baker Lab, bioRxiv 2025)
  • targetdiff - 3D Equivariant Diffusion for Target-Aware Molecule Generation (ICLR2023)
  • SeFMol (Science Advances 2026) - Semi-flexible molecular diffusion model for structure-based drug design with reinforcement learning, achieving 20× faster sampling and providing a no-code web platform for molecular design (ISPC Lab, Tongji University, 2026)
  • ReQFlow - Rectified Quaternion Flow for efficient protein backbone generation, 37× faster than RFDiffusion with 0.972 designability (ICML 2025)
  • AlphaFlow - AlphaFold fine-tuned with flow matching for generating protein conformational ensembles, covering both experimental PDB states and molecular dynamics ensembles at physiological temperatures; includes ESMFlow variant (MIT, 526+ stars, 2024)
  • BioEmu - Microsoft's generative model for sampling protein equilibrium conformations 100,000× faster than MD simulations, predicting domain motions, local unfolding and cryptic binding pockets on a single GPU (Science 2025)
  • STARLING (Holehouse Lab, Nature 2026) - Latent-space probabilistic denoising diffusion model for predicting coarse-grained conformational ensembles of intrinsically disordered proteins and regions from sequence, with GPU/CPU inference, trajectory export, and FAISS-based similarity search (67+ stars, LGPL-3.0)
  • dynamicPDB (AAAI 2025) - Dynamic Protein Data Bank integrating dynamic behaviors and physical properties into protein structures via a new dataset and SE(3) model extension, enabling richer understanding of protein conformational landscapes (Fudan University, 784+ stars)
  • ProteinMPNN - Deep learning-based protein sequence design (inverse folding) from backbone structures, achieving 52.4% sequence recovery vs 32.9% for Rosetta, core tool in modern protein design pipelines (Baker Lab, Science 2022)
  • LigandMPNN - Extension of ProteinMPNN for protein sequence design in the context of small-molecule ligands, metal ions, and nucleic acids, enabling binding site engineering and co-factor redesign (Baker Lab)
  • ColabDesign - Accessible protein design platform via Google Colab integrating AlphaFold2, RoseTTAFold, and ProteinMPNN for de novo hallucination, fixed backbone design, and binder design (Sergey Ovchinnikov, 2022+)
  • BindCraft - Simple and accurate de novo protein binder design pipeline using AlphaFold2 backpropagation, MPNN, and PyRosetta for automated binder discovery (bioRxiv 2024)
  • Genie 2 - Diffusion model for scalable protein structure design with multi-motif scaffolding capabilities, achieving state-of-the-art designability, diversity, and novelty through SE(3)-equivariant attention and massive data augmentation (AlQuraishi Lab, 2024)
  • Genie 3 (AlQuraishi Lab, 2026) - Fast, all-atom SE(3)-equivariant diffusion model for protein design achieving state-of-the-art performance on unconditional generation, motif scaffolding, and binder design while retaining the computational efficiency of equivariant architectures (bioRxiv 2026)
  • Chroma - Generative model for programmable protein design using diffusion modeling, equivariant graph neural networks, and conditional random fields to efficiently sample diverse all-atom structures; supports conditional generation via composable conditioners for substructure, symmetry, shape, and neural-network predictions; validated crystallographically (Generate Biomedicines, Nature 2023)
  • EvoDiff - Discrete diffusion framework for generative protein sequence design over evolutionary-scale databases, supporting unconditional generation, evolutionary-guided conditional design, motif scaffolding, and intrinsically disordered region generation through order-agnostic autoregressive diffusion, enabling sequence-only protein design without structural priors (Microsoft Research, Nature Communications 2024)
  • DISCO - General multimodal protein design framework enabling DNA-encoding of chemistry for programmable enzyme design and diverse protein generation through diffusion-based generative modeling (190+ stars, Apache 2.0, 2026)
  • SwitchCraft - Programmatic framework for designing state-switching proteins via backpropagation through compositional design constraints parameterized by structure prediction models; enables de novo design of allosteric regulators and fluorescent biosensors for arbitrary small-molecule analytes (79+ stars, MIT License, ICML 2026)
  • RFdiffusion3 - Latest RFdiffusion for protein structure design with 10× speedup and atom-level precision (December 2025)
  • RFantibody - Structure-based de novo antibody design pipeline built on RFdiffusion for computational generation of target-specific antibodies (RosettaCommons, 2025)
  • IgGM - Generative foundation model for functional antibody and nanobody design, supporting de novo generation, affinity maturation, inverse design, structure prediction, and humanization (Tencent AI4S, ICLR 2025)
  • DrugAssist - LLM-based molecular optimization tool
  • GenMol - ICML 2025 drug discovery generalist using masked discrete diffusion and fragment-based generation with molecular context guidance (NVIDIA)
  • FoldBack (AISciLab, 2026) - Target-aware peptide design framework that treats receptor sequence and structure as context via multimodal adapter tuning of protein language models (ESMC + ProteinMPNN features), with reinforcement-learning-based 3D dynamic feedback (ESMFold structure evaluation) to suppress unrealistic peptide conformations; ships with a systematic assessment pipeline covering peptide-target affinity, structure quality, physicochemical properties, diversity, and novelty (142+ stars, Apache 2.0, 2026)
  • REINVENT - Industrial-grade reinforcement-learning-based generative platform for de novo molecular design with transformer architectures, supporting multi-objective optimization, scaffold decoration, and curriculum learning (AstraZeneca MolecularAI, REINVENT 4, 2024)
  • mint - Learning the language of protein-protein interactions
  • Mol-Instructions - Large-scale biomolecular instruction dataset for chemistry/biology LLMs (ICLR2024)
  • Uni-Mol - Universal 3D molecular pretraining framework with 209M conformations, scaling to 1.1B parameters (Uni-Mol2) on 800M conformations for molecular property prediction, docking, and quantum chemistry (ICLR 2023, NeurIPS 2024)
  • ChemBERTa - Chemical language model
  • DeepChem - Machine learning for chemistry
  • TorchDrug - Powerful and flexible machine learning platform for drug discovery, providing comprehensive tools for molecular property prediction, generative models, knowledge graph reasoning, and reaction prediction with PyTorch backend (1.5K+ stars)
  • DeepMol - Unified ML/DL framework for drug discovery workflows, integrating RDKit, DeepChem, and scikit-learn with SHAP explainability
  • Chemprop - Message passing neural networks for molecule property prediction, ADMET modeling, and reaction prediction, achieving SOTA on MoleculeNet and widely used in pharmaceutical drug discovery (MIT, 2.3K+ stars)
  • RDKit - Cheminformatics toolkit
  • nvMolKit (NVIDIA BioNeMo, 2025) - High-performance, GPU-accelerated library for key computational chemistry tasks including molecular similarity, conformer generation, and geometry relaxation, designed to accelerate drug-discovery and molecular-modeling workflows (264+ stars, Apache 2.0)
  • Open Targets - Open-source data integration platform for systematic drug target identification and prioritization, combining genetics, genomics, chemistry, and pharmacology data from EMBL-EBI, Wellcome Sanger Institute, and pharmaceutical partners to accelerate therapeutic discovery
  • ESM3 - 98B-parameter frontier generative model jointly reasoning over protein sequence, structure, and function, trained on 2.78 billion proteins; generated a novel fluorescent protein (esmGFP) with only 58% sequence identity to known GFPs (EvolutionaryScale, 2024)
  • ESM Cambrian / ESMC (EvolutionaryScale & Chan Zuckerberg Biohub, arXiv 2025) - Frontier protein language models (300M/600M/6B) trained on billions of protein sequences, establishing a new unsupervised scaling frontier beyond ESM2 with emergent long-range structural understanding; ships with ESMFold2 structure prediction (SOTA DockQ pass-rates on Foldbench protein-protein and antibody-antigen complexes, lab-validated de novo binder/scFv design protocol) and the ESM Atlas mapping 6.8B proteins with sparse-autoencoder-interpretable world-model features (2.9K+ stars, 2025-2026)
  • ProtTrans - State-of-the-art pretrained language models for proteins trained on thousands of GPUs and Google TPUs using Transformer architectures, enabling protein property prediction, feature extraction, and transfer learning across diverse downstream tasks (1.3K+ stars, MIT, 2020-2026)
  • ProGen3 (Profluent, 2025) - Public release of Profluent's ProGen3 protein language model family, including PMC-15B supporting sequence- and structure-conditioned generation for protein design, zero-shot fitness prediction, and antibody engineering with state-of-the-art performance on fitness and docking benchmarks (114+ stars, Apache 2.0)
  • ProstT5 (NAR Genomics and Bioinformatics 2024) - Bilingual protein language model translating between protein sequence and structure, finetuned from ProtT5-XL on 17M AlphaFoldDB structures using Foldseek's 3Di structural alphabet, enabling sequence-to-structure prediction, structure-to-sequence inverse folding, and unified protein representation learning (RostLab, 310+ stars)
  • ESMFold - Protein structure prediction from ESM models
  • SaProt - Structure-aware protein language model using 3D structural vocabulary (Foldseek) for joint sequence-structure pretraining, achieving SOTA on protein engineering and fitness prediction benchmarks (ICML 2024, Westlake University & Repl)
  • InterPLM (Nature Methods 2025) - Discovering interpretable features in protein language models via sparse autoencoders, enabling mechanistic understanding of PLM representations for protein engineering and design (288+ stars, MIT License)
  • AiCE (Cell 2025) - AI-assisted mutation nomination approach optimizing protein function by integrating structural and evolutionary constraints into protein inverse folding models, compatible with ProteinMPNN, LigandMPNN, ESM-IF1, and SaProt (Chinese Academy of Sciences, 359+ stars)
  • EVOLVEpro - In silico directed evolution framework using few-shot active learning to optimize protein activities, enabling rapid protein engineering with minimal experimental data (352+ stars, 2023)
  • DPLM (ByteDance, ICML 2024 / ICLR 2025) - Family of diffusion protein language models demonstrating versatile generative and predictive capabilities for protein sequences and structures, including multimodal co-generation, conditional folding, inverse folding, motif scaffolding, and representation learning, with open pretrained weights and training scripts (327+ stars, ICML 2024, ICLR 2025, ICML 2025 Spotlight)
  • Foldseek - Fast and accurate protein structure search using a learned 3Di structural alphabet (VQ-VAE) that discretizes tertiary interactions into structural tokens, enabling protein-universe-scale structural alignment at sequence-search speeds (4-5 orders of magnitude faster than DALI/TM-align) and underpinning many AI4S tools such as SaProt, ESMAtlas search, and AFDB clustering pipelines (Steinegger Lab, Nature Biotechnology 2023)
  • ImmunoStruct (Nature Machine Intelligence 2025) - Multimodal deep learning framework integrating peptide-MHC protein sequence, structure, and biochemical properties to predict class-I immunogenicity for infectious disease epitopes and cancer neoepitopes with cancer-wildtype contrastive learning, enabling personalized vaccine design (Krishnaswamy Lab, Yale University)
  • mosaic - Composite-objective protein design framework integrating Boltz, AlphaFold2, OpenFold3, ProteinMPNN, and ESM via JAX-based gradient optimization over continuous relaxed sequence space for multi-property binder design (319+ stars, MIT License, 2025)
  • TxGemma (Google DeepMind, 2025) - Open Gemma-2-based LLM family (2B/9B/27B) for therapeutics development, fine-tuned on 7M examples from Therapeutics Data Commons; supports classification, regression, and generation tasks across small molecules, proteins, nucleic acids, diseases, and cell lines, with chat variants for scientific dialogue and agentic integration (Agentic-Tx), matching or beating SOTA on 64/66 therapeutic tasks (open weights on HuggingFace)

Genomics & Bioinformatics

  • RhoFold+ - End-to-end RNA 3D structure prediction using RNA language model pretrained on 23.7M sequences, outperforming existing methods and human expert groups on RNA-Puzzles and CASP15 (Nature Methods 2024)
  • NuFold (Nature Communications 2025) - End-to-end deep learning approach for RNA tertiary structure prediction with a flexible nucleobase center representation, achieving ~7 Å C1' RMSD across test RNAs and predicting ~545,000 structures covering 2,200+ RNA families (Kihara Lab, Purdue University, 50+ stars)
  • RNA-FM (Nature Methods 2024) - RNA foundation model trained on millions of RNA sequences for generalist RNA sequence understanding, enabling downstream structure prediction, function annotation, and representation learning for non-coding RNAs (ml4bio, 372+ stars)
  • RiNALMo (Nature Communications 2025) - General-purpose RNA language model with 650M parameters pretrained on 36M non-coding RNA sequences, achieving strong generalization on structure prediction tasks including secondary structure prediction, splice-site prediction, mean ribosome loading, and ncRNA classification (lbcb-sci, 165+ stars, Apache-2.0)
  • RNAPro (NVIDIA, 2026) - State-of-the-art RNA 3D folding model developed with Stanford Das Lab and Kaggle competition winners, featuring a 488M-parameter AF3-like architecture with MSA and template-based modeling, enabling structure-driven drug discovery and RNA therapeutics design (NVIDIA-Digital-Bio, Apache 2.0)
  • gRNAde - Generative AI framework for inverse design of 3D RNA structure and function using geometric deep learning, learning design rules from 3D structures to capture complex tertiary interactions (pseudoknots, non-canonical base pairs) with expert-level accuracy for designing functional RNAs including aptamers and ribozymes (bioRxiv 2025)
  • AIDO.ModelGenerator - GenBio AI's software stack for the AI-Driven Digital Organism, supporting adaptation and finetuning of multiscale biological foundation models across DNA, RNA, protein, structure, and single-cell tasks with reproducible CLIs and pretrained model zoo (2025)
  • Evo 2 - Arc Institute's 40B-parameter genome foundation model trained on 9 trillion nucleotides from all domains of life, supporting 1M base pair context for generalist DNA/RNA/protein prediction and design (Nature 2026)
  • Carbon (Hugging Face, 2026) - Family of causal genomic foundation models trained on 1T tokens (~6T DNA base pairs) from the Carbon Pretraining Corpus, combining eukaryote genes, mRNA transcripts, and prokaryote genomes with a hybrid text/6-mer tokenizer; Carbon-3B matches or beats Evo2-7B on zero-shot DNA evaluations including sequence recovery, variant effect prediction, and perturbations (Apache 2.0, 201+ stars)
  • Nucleotide Transformer - Foundation models for genomics and transcriptomics pretrained on 3,000+ human genomes and 850+ diverse species, enabling chromatin accessibility prediction, splice site detection, and promoter classification across multiple model scales (InstaDeep, NVIDIA & TUM, Nature Methods 2023)
  • HyenaDNA - Long-range genomic foundation model using subquadratic Hyena operators instead of Transformer attention, enabling context lengths up to 1 million nucleotides for chromosome-scale DNA sequence modeling and downstream genomics tasks (Stanford Hazy Research, NeurIPS 2023, 784+ stars, Apache 2.0)
  • Caduceus (ICML 2024) - Bi-directional DNA language model based on the Mamba state space architecture, enabling efficient long-range genomic sequence modeling with linear-time complexity and built-in reverse-complement equivariance; achieves strong performance on chromatin accessibility, enhancer, and promoter prediction benchmarks (Stanford & UC Berkeley, 500+ stars)
  • CodonFM (NVIDIA) - Family of codon-resolution language models trained on 130 million protein-coding sequences from over 20,000 species, enabling cross-species gene expression prediction and codon-level functional genomics (2025)
  • LucaOne - Generalized biological foundation model with unified nucleic acid and protein language, integrating DNA/RNA/protein sequences (Nature Machine Intelligence 2025)
  • Geneformer - Single-cell transformer foundation model pretrained on 104M human transcriptomes via masked gene prediction, enabling transfer learning for cell type classification, gene network analysis, and in silico perturbation with limited labeled data (Nature 2023, V2 2024)
  • Nicheformer - Foundation model jointly trained on single-cell and spatial transcriptomics data, enabling unified representation learning across cellular and tissue spatial contexts for cell type prediction, spatial domain inference, and cross-modal integration (theislab, bioRxiv 2024, 164+ stars)
  • scFoundation - 100M-parameter foundation model pretrained on 50M+ human single-cell transcriptomes covering ~20,000 genes, achieving SOTA on gene expression enhancement, drug response and perturbation prediction (Nature Methods 2024)
  • scPRINT (Nature Communications 2025) - Large transformer-based single-cell foundation model pretrained on 50 million cells for robust gene network inference, expression denoising, cell embedding, and zero-shot label prediction, leveraging ESM2 protein embeddings and bidirectional transformer architecture (Cantini Lab, 148+ stars, GPL-3.0)
  • Tahoe-x1 - Apache 2.0 single-cell foundation model family scaling to 3B parameters, pretrained on 266M cell profiles including perturbation data and released with training, embedding, and downstream benchmarking workflows for disease-relevant single-cell tasks (2025)
  • Stack - Arc Institute's single-cell foundation model enabling in-context learning at inference time via a novel tabular attention architecture, trained on 150M uniformly-preprocessed cells for generalizing biological effects and generating unseen cell profiles in novel contexts (2025)
  • State (Arc Institute, bioRxiv 2025) - Machine learning model predicting cellular perturbation response across diverse contexts with State Transition (ST) and State Embedding (SE) variants, featuring CLI tooling, PyPI distribution, and Virtual Cell Challenge integration (575+ stars)
  • TranscriptFormer (Chan Zuckerberg Initiative, bioRxiv 2025) - Family of generative single-cell foundation models (TF-Metazoa, TF-Exemplar, TF-Sapiens) jointly modeling genes and their expression levels via expression-aware autoregressive transformers, trained on up to 112M cells across 12 species spanning 1.53 billion years of evolution; achieves robust zero-shot cell type classification across species, disease state identification in human cells, and prediction of cell-type-specific transcription factors and gene-gene regulatory relationships, pip-installable with pretrained weights (CZI, 166+ stars, MIT License)
  • scvi-tools - Deep probabilistic framework for single-cell and spatial omics analysis, integrating scVI, scANVI, totalVI and other VAE-based models for batch correction, cell annotation, multi-omics integration, and RNA velocity (scverse/NumFOCUS, Nature Methods 2018/2024)
  • CellRank - Probabilistic framework for inferring cell fate decisions and trajectory dynamics from multi-view single-cell data using Markov chains and machine learning, integrating RNA velocity, pseudotime, and metabolic labeling to predict differentiation paths and terminal states (scverse/Theis Lab, 449+ stars, BSD 3-Clause)
  • cellxgene (Chan Zuckerberg Initiative) - Interactive explorer for single-cell transcriptomics data enabling visualization of UMAP/t-SNE embeddings, differential expression analysis, and cross-dataset comparison through a fast web-based interface; widely adopted for exploring atlas-scale single-cell datasets and integrating with AI/ML analysis workflows (773+ stars, MIT License)
  • Helical - Unified framework for state-of-the-art pre-trained bio foundation models across genomics and transcriptomics, providing standardized interfaces and pipelines for DNA, RNA, and single-cell models including Evo 2, Geneformer, scGPT, and UCE with streamlined inference, benchmarking, and fine-tuning workflows (213+ stars, 2024-2025)
  • GEARS - Geometric deep learning model predicting transcriptional outcomes of novel single- and multi-gene perturbations using gene–gene knowledge graphs, 40% higher precision than prior methods on combinatorial perturbation prediction (Stanford, Nature Biotechnology 2024)
  • scDFM (ICLR 2026) - Distributional flow matching model for robust single-cell perturbation prediction, modeling the full distribution of perturbed cellular expression profiles conditioned on control states via PAD-Transformer and multi-kernel MMD regularization; reduces MSE by 19.6% over the strongest baseline in combinatorial settings (Westlake University, 41+ stars, MIT License)
  • scTranslator (Nature Biomedical Engineering 2025) - Pre-trained large generative model translating single-cell transcriptomes to proteomes in an alignment-free manner, generating absent protein abundance data for CITE-seq, spatial CITE-seq, REAP-seq, and NEAT-seq across tissues and diseases; offers three model variants pretrained on 2M human cells, 160K PBMCs, or 18K bulk samples (Tencent AI Lab Healthcare, 96+ stars)
  • scGPT - Single-cell analysis with transformers
  • UCE (Stanford SNAP, Nature 2024) - Universal Cell Embeddings: zero-shot single-cell foundation model pretrained on 36M cells across 11M species, learning cross-species gene function representations via a protein-language-model-informed token space; enables zero-shot cell type annotation, embedding, and integration of unseen datasets and species without fine-tuning (338+ stars, MIT License)
  • CellWhisperer (Nature Biotechnology 2025) - Multimodal AI bridging transcriptomics data and natural language, enabling intuitive chat-based exploration and analysis of single-cell RNA-seq datasets through conversational interaction without coding; fine-tuned Mistral 7B LLaVA model emulating biologist-bioinformatician discussions (207+ stars, GPL-3.0)
  • CellTypist - Automated cell type annotation tool for single-cell transcriptomics using gradient boosting and logistic regression with reference atlases, enabling standardized classification across datasets (Wellcome Sanger Institute, Nature Biotechnology 2022)
  • mLLMCelltype - Multi-LLM consensus framework for automated cell type annotation in single-cell transcriptomics, integrating predictions from 10+ large language models with iterative discussion and uncertainty quantification to reduce single-model biases, achieving up to 95% accuracy without reference datasets; available as CRAN R package and PyPI Python package with Scanpy/Seurat integration (2025)
  • Cell2Sentence - Teaching Large Language Models the Language of Biology through single-cell transcriptomics (ICML 2024)
  • OmicVerse - Unified Python framework for bulk, single-cell, and spatial RNA-seq multi-omics analysis with deep learning deconvolution (VAE) and graph neural networks, bridging Bindea, Bindea, scanpy and squidpy ecosystems (Nature Communications 2024)
  • ChatSpatial - MCP server enabling spatial transcriptomics analysis via natural language, integrating 60+ methods including SpaGCN, Cell2location, LIANA+, CellRank for Visium, Xenium, MERFISH platforms
  • Enformer - Gene expression prediction
  • DNABERT - DNA sequence analysis
  • DNABERT-2 (ICLR 2024) - Efficient foundation model and benchmark for multi-species genome understanding with context-aware nucleotide representations, improving upon DNABERT for diverse genomic task transfer learning (UIUC MAGICS Lab, 484+ stars)
  • gReLU (Genentech, 2024) - Python library to train, interpret, and apply deep learning models to DNA sequences, providing a unified framework for regulatory genomics with support for CNN and transformer architectures, variant effect prediction, and attribution analysis (325+ stars)
  • BioReason (NeurIPS 2025) - First architecture deeply integrating a DNA foundation model with an LLM for multimodal biological reasoning, achieving 98% accuracy on KEGG disease pathway prediction and 15%+ average gains on variant effect prediction with interpretable step-by-step reasoning traces (bowang-lab, 390+ stars)
  • scBERT - Single-cell BERT for gene expression
  • GenePT - Generative pre-training for genomics
  • DNA Claude Analysis - Interactive personal genome analysis toolkit using Claude Code and Python. Parses raw genotyping data from consumer DNA services and analyzes SNPs across 17 categories including health risks, pharmacogenomics, ancestry, and nutrition, with a terminal-style HTML dashboard.
  • OpenCRISPR - First open-source AI-generated gene editing systems developed with protein language models, enabling programmable CRISPR-Cas nucleases for synthetic biology and therapeutic genome editing (Profluent, 2024)
  • AlphaMissense - Google DeepMind's AlphaFold-derived classifier for proteome-wide missense variant effect prediction, providing pathogenicity scores for all ~71M possible human missense variants and classifying 89% with 90% precision; pre-computed predictions are integrated into Ensembl VEP and UCSC Genome Browser to support clinical variant interpretation (Science 2023)
  • AlphaGenome - Google DeepMind's unified DNA sequence foundation model predicting molecular consequences of genetic variants from single-base resolution up to 1 megabase context, jointly outputting thousands of regulatory tracks (RNA expression, splicing, chromatin accessibility, TF binding, contact maps) for human and mouse genomes via a Python client and non-commercial API (2025)
  • GPN-Star (Song Lab, UC Berkeley, bioRxiv 2025) - Phylogeny-aware genomic language model trained on whole-genome alignments across multiple evolutionary timescales, predicting functional constraints and variant effects for human, mouse, chicken, fly, worm, and Arabidopsis genomes (344+ stars, MIT License)
  • GENERanno (bioRxiv 2025) - Genomic foundation model for metagenomic and genome annotation, featuring an 8k base-pair context and 500M parameters trained on 386B base pairs of eukaryotic DNA; provides expert models and a unified CLI for prokaryotic/eukaryotic coding-sequence annotation with strong performance on Genomic Benchmarks, Nucleotide Transformer tasks, and custom Gener tasks (GenerTeam, 314+ stars, MIT License)
  • GENERator (bioRxiv 2026) - Long-context generative genomic foundation model using 6-mer tokenization for DNA sequence modeling and generation, with v2 model families for prokaryote and eukaryote genomes and pretrained weights available on HuggingFace (GenerTeam, 460+ stars, MIT License, 2025-2026)
  • DeepVariant - Google DeepMind's deep learning analysis pipeline for calling genetic variants (SNPs and indels) from next-generation DNA sequencing data, achieving human expert-level accuracy and widely adopted in clinical genomics, population genetics, and precision medicine; pre-trained models available for multiple sequencing platforms and organismal genomes (Nature Biotechnology 2018, 3.7K+ stars)
  • Casanovo - Transformer encoder-decoder for de novo peptide sequencing from tandem mass spectrometry, translating MS/MS spectra directly to peptide sequences without reference databases, enabling identification of novel peptides for immunopeptidomics, antibody repertoires, and metaproteomes (Noble Lab UW, Nature Communications 2024)
  • InstaNovo (InstaDeep, Nature Machine Intelligence 2025) - Transformer that translates fragment ion peaks into peptide sequences for database-free de novo sequencing in large-scale proteomics, with InstaNovo+ extending it as a multinomial diffusion model that iteratively refines predicted sequences, plus InstaNovo-P for phosphoproteomics and Winnow for calibrated confidence with FDR control (130+ stars, Apache 2.0, actively maintained)
  • DreaMS (Nature Biotechnology 2025) - Transformer foundation model for tandem mass spectrometry (MS/MS) self-supervised on millions of unannotated spectra from the GeMS dataset via masked peak prediction and chromatographic retention-order objectives, producing 1024-dimensional molecular representations; achieves SOTA on spectral similarity, chemical property, and molecular fingerprint prediction, and powers the DreaMS Atlas annotating 201M+ MS/MS spectra for metabolomics and natural product discovery (Pluskal Lab, IOCB Prague & MIT, 211+ stars, MIT License)
  • Dorado - Oxford Nanopore's official deep-learning basecaller for nanopore sequencing, converting raw electrical signals into DNA/RNA sequences with integrated modified-base (methylation) detection and efficient CPU/GPU inference; foundational tool for long-read genomics, epigenetics, and real-time sequencing analysis (nanoporetech, 846+ stars, actively maintained)
  • geNomad (Nature Biotechnology 2023) - Hybrid deep learning and alignment-based tool for identifying viruses, plasmids, and other mobile genetic elements in isolates, metagenomes, and metatranscriptomes, combining neural-network gene-content classifiers with nucleotide-sequence signatures; also performs viral taxonomic assignment, provirus detection in host