{"slug": "eliminating-conda-cuda-dependency-hell-in-computational-biology-pipelines", "title": "Eliminating Conda/CUDA dependency hell in computational biology pipelines", "summary": "Evo Design released proto-tools, an open-source Python package that provides a unified interface for computational biology and biological AI tools, eliminating dependency conflicts by running each tool in an automatically managed isolated environment. The package supports language models, structure predictors, inverse folding, sequence analysis, and more, and is available under an MIT license via pip install from GitHub.", "body_md": "Welcome! This repository contains the open-source implementation of `proto-tools`\n\n, a Python package containing a large suite of computational biology and biological AI tools, all accessible through a single, consistent Python interface. Language models, structure predictors, inverse folding, sequence analysis, gene annotation, conformational dynamics, genomic scoring, and more are all available through a single `pip install`\n\ncommand.\n\nEvery tool runs in its own automatically managed isolated environment, so all dependency wrangling is handled for you. In addition, `proto-tools`\n\nimplements extensive infrastructure for features such as device management and GPU fan-out, making it easy to call tools in quick succession. You can use it as a standalone Python library, as part of the broader [proto-language](https://github.com/evo-design/proto-language) optimization system, or through the [proto-client](https://github.com/evo-design/proto-client) Python SDK for hosted access over the Proto Bio API.\n\nProto-tools is open source under an MIT license. Contributions are welcome!\n\nProto-tools requires Python 3.10+:\n\n```\npip install git+https://github.com/evo-design/proto-tools.git\n```\n\nNote\n\nA direct PyPI install (`pip install proto-tools`\n\n) will be available soon.\n\nNote\n\nIf you are developing or contributing to this project, follow the setup instructions in [CONTRIBUTING.md](/evo-design/proto-tools/blob/main/CONTRIBUTING.md) instead.\n\nAll persistent data (model weights and tool environments) is cached under the `PROTO_HOME`\n\ndirectory on first use (defaults to `~/.proto/`\n\n).\n\nTo customize the storage location, you can specify a path via the following environment variable:\n\n```\n# Add to your shell profile:\nexport PROTO_HOME=/path/to/your/proto_home\n```\n\nFor shared filesystems, model weights can be reused to avoid downloading duplicate copies. The `PROTO_MODEL_CACHE`\n\nenvironment variable lets you point just the weights at that shared location (sharing tool environments is not recommended): `export PROTO_MODEL_CACHE=/path/to/shared/weights`\n\n. See [notes/storage.md](/evo-design/proto-tools/blob/main/notes/storage.md) for all details and options.\n\nA few tools use gated models or software that require accepting a license / terms-of-use first (e.g. ESM3, AlphaGenome, AlphaFold3, X3DNA). See [notes/gated-models.md](/evo-design/proto-tools/blob/main/notes/gated-models.md) for the full list and per-model access steps.\n\nTip\n\n**You're all set up!** To learn what features are available in the library, check out the [guides](/evo-design/proto-tools/blob/main/guides) — four short notebooks covering tool environments, persistent execution, device management, and parallel multi-GPU runs.\n\n```\nbinder_design/                  # De novo antibody / binder design pipelines\n├── bindcraft/\n├── freebindcraft/\n└── germinal/\ncausal_models/                 # Autoregressive sequence models\n├── evo1/\n├── evo2/\n├── progen2/\n└── progen3/\ndatabase_retrieval/             # Sequence and structure database access\n├── alphafold_db/\n├── alphamissense_db/\n├── ccd_lookup/\n├── ensembl/\n├── interproscan/\n├── ncbi/\n├── pdb/\n├── pubchem/\n├── sequence_fetch/\n└── uniprot/\ngene_annotation/                # Sequence annotation\n├── crispr_tracr_rna/\n├── meme/\n├── minced/\n├── miranda/\n├── promoter_calculator/\n└── pyhmmer/\ninverse_folding/                # Sequence design from structures\n├── esm_if1/\n├── fampnn/\n├── ligandmpnn/\n└── proteinmpnn/\nmasked_models/                  # Masked language models\n├── ablang/\n├── esm2/\n├── esm3/\n└── esmc/\nmutagenesis/                    # Random sequence mutagenesis\n├── random_nucleotide/\n└── random_protein/\norf_prediction/                 # Open reading frame detection\n├── orfipy/\n└── prodigal/\nrna_splicing/                   # RNA splice site prediction\n├── pangolin/\n├── splice_transformer/\n└── spliceai/\nsequence_alignment/             # Sequence search and multiple sequence alignment\n├── blast/\n├── mafft/\n└── mmseqs2/\nsequence_scoring/               # Genomic and regulatory scoring\n├── alphagenome/\n├── borzoi/\n├── deeppbs_specificity/\n├── enformer/\n├── malinois/\n├── na_mpnn_specificity/\n├── puffin/\n└── segmasker/\nstructure_alignment/            # Structure comparison\n├── foldmason/\n├── foldseek/\n├── pymol_rmsd/\n├── tmalign/\n└── usalign/\nstructure_design/               # De novo structure generation\n└── rfdiffusion3/\nstructure_dynamics/             # Conformational dynamics\n└── bioemu/\nstructure_prediction/           # 3D structure prediction\n├── alphafold2/\n├── alphafold3/\n├── boltz2/\n├── chai1/\n├── esmfold/\n├── esmfold2/\n├── protenix/\n├── rf3/\n├── viennarna/\n└── x3dna/\nstructure_scoring/              # Structure quality scoring\n├── dssp/\n├── ipsae/\n├── metal3d/\n├── pdockq2/\n├── pyrosetta/\n└── structure_metrics/\n```\n\nRunnable walkthroughs of the core framework features live in [ guides/](/evo-design/proto-tools/blob/main/guides) and are also available on our\n\n[docs page](https://proto.evodesign.org/docs/tools/introduction):\n\n[Tool Environments](/evo-design/proto-tools/blob/main/guides/tool_environments.ipynb)— how isolated environments are built and cached on first call.[Tool Persistence](/evo-design/proto-tools/blob/main/guides/tool_persistence.ipynb)— keep models warm across calls[Device Management](/evo-design/proto-tools/blob/main/guides/device_management.ipynb)— GPU allocation, LRU eviction, CPU offload[Parallel Execution](/evo-design/proto-tools/blob/main/guides/parallel_execution.ipynb)— fan out work across every GPU with`ToolPool`\n\nEach specific tool also ships a minimal `examples/example.ipynb`\n\nunder `proto_tools/tools/{category}/{tool}/examples/`\n\n.\n\nRun tools through natural language with any coding agent (Claude Code, Gemini CLI, OpenAI Codex CLI, etc.). Point the agent at `proto-tools agent-context`\n\n: it prints a primer covering the `Input → Config → run_*() → Output`\n\npattern, the offline CLI discovery verbs, persistence and parallel execution, and links to the long-form notes on GitHub. The command ships in the wheel, so it works on a plain `pip install`\n\nwith no repo checkout.\n\nIf you've cloned the repo for contributing, agents also pick up `CLAUDE.md`\n\n(symlinked as `AGENTS.md`\n\n/`GEMINI.md`\n\n) and the task-specific guides in [ .claude/skills/](/evo-design/proto-tools/blob/main/.claude/skills) automatically.\n\nSee [CONTRIBUTING.md](/evo-design/proto-tools/blob/main/CONTRIBUTING.md) for full developer setup, storage\nconfiguration, PR format, code style, and testing conventions.\n\nIf you use Proto in your research, please cite our preprint:\n\nMerchant AT, Guo D, Viggiano B, Brennan-Almaraz LE, Hur E, Mai T, Yin P, King SH, Ashley E, Hie BL.\n\nA high-level programming language for generative biology with Proto.bioRxiv(2026). doi:[10.64898/2026.06.22.733870]\n\n```\n@article{Merchant2026.06.22.733870,\n  author = {Merchant, Aditi T and Guo, Daniel and Viggiano, Ben and Brennan-Almaraz, Lucas Emmanuel and Hur, Evelyn and Mai, Tina and Yin, Peter and King, Samuel H and Ashley, Euan and Hie, Brian L},\n  title = {A high-level programming language for generative biology with Proto},\n  elocation-id = {2026.06.22.733870},\n  year = {2026},\n  doi = {10.64898/2026.06.22.733870},\n  publisher = {Cold Spring Harbor Laboratory},\n  URL = {https://www.biorxiv.org/content/10.64898/2026.06.22.733870},\n  journal = {bioRxiv}\n}\n```\n\n", "url": "https://wpnews.pro/news/eliminating-conda-cuda-dependency-hell-in-computational-biology-pipelines", "canonical_source": "https://github.com/evo-design/proto-tools", "published_at": "2026-07-21 04:27:27+00:00", "updated_at": "2026-07-21 04:53:08.740866+00:00", "lang": "en", "topics": ["artificial-intelligence", "developer-tools", "machine-learning"], "entities": ["Evo Design", "proto-tools", "proto-language", "proto-client", "Proto Bio API", "ESM3", "AlphaGenome", "AlphaFold3"], "alternates": {"html": "https://wpnews.pro/news/eliminating-conda-cuda-dependency-hell-in-computational-biology-pipelines", "markdown": "https://wpnews.pro/news/eliminating-conda-cuda-dependency-hell-in-computational-biology-pipelines.md", "text": "https://wpnews.pro/news/eliminating-conda-cuda-dependency-hell-in-computational-biology-pipelines.txt", "jsonld": "https://wpnews.pro/news/eliminating-conda-cuda-dependency-hell-in-computational-biology-pipelines.jsonld"}}