Find open-source science resources

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)

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)

Interaction Fingerprints for protein-ligand complexes and more.

The Zebrafish Activity Prediction Benchmark for forecasting cellular-resolution neural activity throughout an entire vertebrate brain, combining light-sheet microscopy calcium-imaging data, forecasting tasks, and evaluation tools to advance whole-brain neural dynamics modeling (77+ stars, Apache 2.0)

Next-generation benchmark for data-driven global weather models with standardized evaluation framework and curated datasets for ML forecasting (Google Research, 2024)

Trainable PyTorch reproduction of AlphaFold 3

Medical large vision-language model unifying comprehension and generation via heterogeneous knowledge adaptation, enabling holistic medical image understanding, visual question answering, and clinical report generation across diverse modalities (ZJU4HealthCare, 1.6K+ stars)

SQUARNA is a tool for RNA secondary structure prediction. It can take a single RNA sequence or an alignment of sequences as input. SQUARNA handles pseudoknots and can predict alternative structures. SQUARNA allows structural restraints and chemical probing data as additional input and is available at https://github.com/febos/SQUARNA and https://larnal.imol.institute/.

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)

PyTorch-based embedding instance segmentation algorithm optimized for accurate, efficient, and portable cell and nucleus segmentation across fluorescence and brightfield microscopy images, achieving state-of-the-art speed and accuracy with lightweight model sizes suitable for edge deployment (224+ stars, Apache 2.0)

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)

This tool estimates the completeness of KEGG pathway modules from the presence or absence of KEGG orthologues (KOs)

PMGen (Peptide MHC Generator) is a comprehensive pipeline for predicting peptide-MHC (pMHC) complex structures and designing optimized peptide sequences.

Scientific taste learning framework showing AI can judge and propose research ideas with long-term impact; trains Scientific Judge as a generative reward model and Scientific Thinker as an ideation policy using Reinforcement Learning from Community Feedback (RLCF) on large-scale citation signals, with SciJudgeBench and released HuggingFace model weights (425+ stars, Apache 2.0, 2026)

Advanced OCR with PP-StructureV3 document parsing, 13% accuracy improvement, supports 80+ languages

SDK & library for AI-driven scientific computing applications

200+ AI for Science papers with Chinese interpretations

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)

SMBGC Annotation using Neural Networks Trained on Interpro Signatures

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)

A python-based workflow manager.

Generalized Attribute Based Ratings Information Extraction Library; official OpenAI toolkit that turns messy qualitative corpora into analysis-ready datasets for social scientists and data scientists, measuring quantitative attributes in text, images, or audio using the GPT API. See the [official blog post](https://openai.com/index/scaling-social-science-research/) and [NBER working paper](http://www.nber.org/papers/w34834) (413+ stars, Apache 2.0)

Py-HLA-Match is a Python library for standardised, rule-based HLA (Human Leukocyte Antigen) matching in retrospective analyses, method development, benchmarking, and in-silico studies in immunogenetics and related fields.

AI-powered field boundary delineation toolkit combining satellite foundation models, embeddings, and global training data for accurate agricultural parcel/field boundary mapping, with Google Earth Engine integration and PyPI distribution (84+ stars, Apache 2.0, 2026)

Autonomous AI agent for end-to-end spatial proteomics analysis, featuring SP-Bench for agentic multiplexed-imaging workflows (tomtommyyuan, 140+ stars, 2026)

Computational fluid dynamics in JAX, enabling differentiable Navier-Stokes simulations with automatic differentiation for ML-accelerated CFD research, supporting turbulence modeling, convection-diffusion, and complex boundary conditions on CPUs and GPUs (Google Research, 947+ stars)

Multi-agent system for drug-discovery gene target validation. LangGraph agents over an MCP data layer (~26 data sources, ~44 tools) score evidence across six independent lenses (genetics, biology, safety, clinical, commercial, regulatory) into a provenanced dossier. Configurable local/cloud LLM routing with full Langfuse/OTEL traceability.

Democratizing AlphaFold3: PyTorch reimplementation to accelerate protein structure prediction research

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)

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)

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

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)

This tutorial aims to illustrate the process of extracting structural and dynamical properties from a DNA MD trajectory helical parameters, step by step, using the BioExcel Building Blocks library (biobb). The particular example used is the Drew Dickerson Dodecamer sequence -CGCGAATTCGCG- (PDB code 1BNA). The trajectory used is a 500ns-long MD simulation taken from the BigNASim database (NAFlex_DDD_II entry).

First multi-domain generative foundation model for the natural sciences built on a unified scientific grammar, encoding proteins, antibodies, small molecules, chemical reactions, materials, and their spatial interactions into a shared token vocabulary; enables unified generation, prediction, and design across domains under a purely autoregressive paradigm (134+ stars, Apache 2.0, 2026)

DeepMind's graph neural network for materials exploration, discovering 2.2M new crystal structures (380K most stable) equivalent to 800 years of traditional research, with 520K+ materials dataset open-sourced (Nature 2023)

A Molecular Interaction-Guided Graph Learning Framework for Multi-Omics Cancer Classification

Numerical differential equation solving in JAX

Google's open multimodal medical LLM family built on Gemma 3, including a 4B multimodal model handling medical images (radiology, pathology, dermatology) alongside text and a 27B text model for clinical reasoning; trained on de-identified medical data with checkpoints and inference code released under Apache 2.0 (1.6K+ stars)

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)

Gene expression prediction

Universal foundation model for grounded biomedical image interpretation, enabling comprehensive visual understanding, reasoning, and grounding across diverse biomedical imaging modalities with strong zero-shot generalization (55+ stars, Apache 2.0, 2025-2026)

Programmatic data labeling and weak supervision

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)

Surprisingly simple and efficient frontier probabilistic weather forecaster built on a standard U-Net trained with deterministic MAE pre-training followed by short CRPS fine-tuning via Monte Carlo Dropout, matching or exceeding the probabilistic skill of GenCast and IFS ENS at 1.5° resolution with >10× less training compute than leading CRPS models and >10× lower inference latency than diffusion models; trains in under 12 H200 GPU-days and generates a 15-day ensemble forecast in 3 seconds, with official code and pretrained checkpoints (UCLA, 41+ stars, Apache 2.0)

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)

Foundation model for universal cell segmentation achieving state-of-the-art performance across bacteria, tissue, yeast, cell culture, and diverse imaging modalities (brightfield, fluorescence, phase), with pip-installable inference and Napari plugin (vanvalenlab/Caltech, bioRxiv 2024)

Universal time series forecasting via the UNI2TS library, training a single transformer with shared self-attention and specialized mixture-of-experts feed-forward blocks to achieve strong zero-shot generalization across heterogeneous domains including energy, weather, transportation, and health time series (1.6K+ stars, Apache 2.0, 2024-2026)

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)

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)

Generalized biological foundation model with unified nucleic acid and protein language, integrating DNA/RNA/protein sequences (Nature Machine Intelligence 2025)