Find open-source science resources

Microsoft's AI-powered ab initio biomolecular dynamics simulation achieving quantum-mechanical accuracy for proteins with 10,000+ atoms, orders of magnitude faster than DFT using protein fragmentation and ML force fields (Nature 2024)

Equivariant graph attention Transformer (ICLR2023)

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)

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)

Large-scale biomolecular instruction dataset for chemistry/biology LLMs (ICLR2024)

Tools for adding mutations to existing `.bam` files, used for testing mutation callers.

Batteries included genomic analysis pipeline for variant and RNA-Seq analysis, structural variant calling, annotation, and prediction.

Materials informatics benchmark

Structure-aware prefix adaptation for integrating LLMs with knowledge graphs (ACM MM 2024)

Resources on ChIP-seq data which include papers, methods, links to software, and analysis.

UNIX-style FASTA manipulation tools.

Biomedical text generation

A benchmarking platform for molecular generation models.

Ujjwal Karn.

Open language model for mathematics (7B/34B) trained on Proof-Pile-2, outperforming Minerva at equal scale on MATH benchmark, with tool use and formal theorem proving in Lean without finetuning (EleutherAI, ICLR 2024)

A package for benchmarking of models for _de novo_ molecular design.

PINN research collection

OpenChem is a deep learning toolkit for Computational Chemistry with PyTorch backend.

Secure text-to-visualization through standardized chart specifications

First foundation model for weather and climate by Microsoft, Vision Transformer-based architecture trained on heterogeneous datasets (ICML 2023)

Screen a bacterial assembly (contigs/CDS or proteins) for nucleotide or protein sequences. Pipeline that screens for presence of genes of interest (GOI) in bacterial assemblies. Generates multiple CSVs and plots that describe which genes are present and how variable their sequence is. Can use DNA or protein query sequences (GOIs) and DNA contigs/fastas or protein fastas as database (db) to search in.

An open, extensible Python framework for GPU-accelerated alchemical free energy calculations.

Easily submitting PBS jobs with script template. Multiple input files supported.

A Library for Deep Learning in Biology and Chemistry.

A deep learning framework (based on Chainer) with applications in Biology and Chemistry.

Enables machine learning on three-dimensional molecular structure.

a robust molecular representation learning framework against distribution shifts.

Go Get Data; A command line interface for obtaining genomic data.

A cookiecutter template for bioinformatics projects, with a focus on building bioinformatics workflows that can run on the MPI-IE cluster according to FAIR principles.

Hierarchical Generation of Molecular Graphs using Structural Motifs.

Spherical CNNs for astronomy

[@crazyhottommy](https://github.com/crazyhottommy)'s notes on various steps and considerations when doing RNA-seq analysis.

Automated strain separation of low-complexity metagenomes

Crystal property prediction

Computation Pipeline library for python widely used in science and bioinformatics.

Pythonic access to the UCSC Genome database.

NanoSV is a software package that can be used to identify structural genomic variations in long-read sequencing data, such as data produced by Oxford Nanopore Technologies’ MinION, GridION or PromethION instruments, or Pacific Biosciences RSII or Sequel sequencers.

Automatic Filtering, Trimming, Error Removing and Quality Control for fastq data.

Molecule validation and standardization based on [RDKit](http://www.rdkit.org/).

A port of [pyVCF](https://github.com/jamescasbon/PyVCF) using Cython for speed.

Tool to generate a count matrix for expression data in Galaxy. generate_count_matrix reads in one or more input text files with expression counts and produces a single combined file. Each input will have a column in the matrix containing expression values. The column containing gene (or feature) names should be identical for all input count files.

Membrane Protein-Lipid Interaction Database. A large-scale experimentally validated dataset of 80685 residue-level lipid contact annotations across 4712 membrane proteins derived from PDB crystal and cryo-EM structures. Provides pre-computed binary contact labels, continuous distance values, sequence-identity-based cluster assignments, and ready-made train-validation-test splits for machine learning.

Python package for biodatafuse project.

AmsterdamUMCdb is a database of de-identified health data related to tens of thousands of intensive care unit admissions, including demographics, vital signs, laboratory tests and medications.

Reactr is an modularized, Snakemake workflow for automated, species-agnostic characterization of gene families from sequence to experimental design. Given a query protein sequence and NCBI taxonomy IDs (or RefSeq assembly accessions), reactr retrieves genomic data and runs comprehensive analysis across 4 integrated tiers: (1) evolutionary analysis, including homolog detection, domain-based clustering, multiple sequence alignment, and phylogenetic inference; (2) synteny and selection analysis, detecting collinear blocks and calculating Ka/Ks ratios; (3) structural and regulatory characterization, including motif discovery, chromosomal mapping, biochemical property prediction, subcellular localization prediction, and promoter analysis; and (4) experimental design tools, generating PCR primers and scored CRISPR gRNAs for lab validation. Reactr bridges computational prediction and experimental validation, thus enabling rapid transition from genomic discovery to functional studies.