{"id":"a9c24eee-34b2-4530-b0ee-1e0f206a9a59","shortId":"dTB6Px","kind":"skill","title":"boltz","tagline":"Structure prediction using Boltz-1/Boltz-2, an open biomolecular structure predictor. Use this skill when: (1) Predicting protein complex structures, (2) Validating designed binders, (3) Need open-source alternative to AF2, (4) Predicting protein-ligand complexes, (5) Using local","description":"# Boltz Structure Prediction\n\n## Prerequisites\n\n| Requirement | Minimum | Recommended |\n|-------------|---------|-------------|\n| Python | 3.10+ | 3.11 |\n| CUDA | 12.0+ | 12.1+ |\n| GPU VRAM | 24GB | 48GB (L40S) |\n| RAM | 32GB | 64GB |\n\n## How to run\n\n> **First time?** See [Installation Guide](../../docs/installation.md) to set up Modal and biomodals.\n\n### Option 1: Modal\n```bash\ncd biomodals\nmodal run modal_boltz.py \\\n  --input-faa complex.fasta \\\n  --out-dir predictions/\n```\n\n**GPU**: L40S (48GB) | **Timeout**: 1800s default\n\n### Option 2: Local installation\n```bash\npip install boltz\n\nboltz predict \\\n  --fasta complex.fasta \\\n  --output predictions/\n```\n\n## Key parameters\n\n| Parameter | Default | Range | Description |\n|-----------|---------|-------|-------------|\n| `--recycling_steps` | 3 | 1-10 | Recycling iterations |\n| `--sampling_steps` | 200 | 50-500 | Diffusion steps |\n| `--use_msa_server` | true | bool | Use MSA server |\n\n## FASTA Format\n\n```\n>protein_A\nMKTAYIAKQRQISFVK...\n>protein_B\nMVLSPADKTNVKAAWG...\n```\n\n## Output format\n\n```\npredictions/\n├── model_0.cif       # Best model (CIF format)\n├── confidence.json   # pLDDT, pTM, ipTM\n└── pae.npy          # PAE matrix\n```\n\n**Note**: Boltz outputs CIF format. Convert to PDB if needed:\n```python\nfrom Bio.PDB import MMCIFParser, PDBIO\nparser = MMCIFParser()\nstructure = parser.get_structure(\"model\", \"model_0.cif\")\nio = PDBIO()\nio.set_structure(structure)\nio.save(\"model_0.pdb\")\n```\n\n## Comparison\n\n| Feature | Boltz-1 | Boltz-2 | AF2-Multimer |\n|---------|---------|---------|--------------|\n| MSA-free mode | Yes | Yes | No |\n| Diffusion | Yes | Yes | No |\n| Speed | Fast | Faster | Slower |\n| Open source | Yes | Yes | Yes |\n\n## Sample output\n\n### Successful run\n```\n$ boltz predict --fasta complex.fasta --output predictions/\n[INFO] Loading Boltz-1 weights...\n[INFO] Predicting structure...\n[INFO] Saved model to predictions/model_0.cif\n\npredictions/confidence.json:\n{\n  \"ptm\": 0.78,\n  \"iptm\": 0.65,\n  \"plddt\": 0.81\n}\n```\n\n**What good output looks like:**\n- pTM: > 0.7 (confident global structure)\n- ipTM: > 0.5 (confident interface)\n- pLDDT: > 0.7 (confident per-residue)\n- CIF file: ~100-500 KB for typical complex\n\n## Decision tree\n\n```\nShould I use Boltz?\n│\n├─ What are you predicting?\n│  ├─ Protein-protein complex → Boltz ✓ or Chai or ColabFold\n│  ├─ Protein + ligand → Boltz ✓ or Chai\n│  └─ Single protein → Use ESMFold (faster)\n│\n├─ Need MSA?\n│  ├─ No / want speed → Boltz ✓\n│  └─ Yes / maximum accuracy → ColabFold\n│\n└─ Why Boltz over Chai?\n   ├─ Open weights preference → Boltz ✓\n   ├─ Boltz-2 speed → Boltz ✓\n   └─ DNA/RNA support → Consider Chai\n```\n\n## Typical performance\n\n| Campaign Size | Time (L40S) | Cost (Modal) | Notes |\n|---------------|-------------|--------------|-------|\n| 100 complexes | 30-45 min | ~$8 | Standard validation |\n| 500 complexes | 2-3h | ~$35 | Large campaign |\n| 1000 complexes | 4-6h | ~$70 | Comprehensive |\n\n**Per-complex**: ~15-30s for typical binder-target complex.\n\n---\n\n## Verify\n\n```bash\nfind predictions -name \"*.cif\" | wc -l  # Should match input count\n```\n\n---\n\n## Troubleshooting\n\n**Low confidence**: Increase recycling_steps\n**OOM errors**: Use MSA-free mode or A100-80GB\n**Slow prediction**: Reduce sampling_steps\n\n### Error interpretation\n\n| Error | Cause | Fix |\n|-------|-------|-----|\n| `RuntimeError: CUDA out of memory` | Complex too large | Use `--use_msa_server false` or larger GPU |\n| `KeyError: 'iptm'` | Single chain only | Ensure FASTA has 2+ chains |\n| `FileNotFoundError: weights` | Missing model | Run `boltz download` first |\n| `ValueError: invalid residue` | Non-standard AA | Check for modified residues in sequence |\n\n### Boltz-1 vs Boltz-2\n\n| Aspect | Boltz-1 | Boltz-2 |\n|--------|---------|---------|\n| Speed | Fast | ~2x faster |\n| Accuracy | Good | Improved |\n| Ligands | Basic | Better support |\n| Release | 2024 | Late 2024 |\n\n---\n\n**Next**: `protein-qc` for filtering and ranking.","tags":["boltz","protein","design","skills","adaptyvbio","agent-skills","claude-code","protein-design","protein-engineering"],"capabilities":["skill","source-adaptyvbio","skill-boltz","topic-agent-skills","topic-claude-code","topic-protein-design","topic-protein-engineering"],"categories":["protein-design-skills"],"synonyms":[],"warnings":[],"endpointUrl":"https://skills.sh/adaptyvbio/protein-design-skills/boltz","protocol":"skill","transport":"skills-sh","auth":{"type":"none","details":{"cli":"npx skills add adaptyvbio/protein-design-skills","source_repo":"https://github.com/adaptyvbio/protein-design-skills","install_from":"skills.sh"}},"qualityScore":"0.513","qualityRationale":"deterministic score 0.51 from registry signals: · indexed on github topic:agent-skills · 126 github stars · SKILL.md body (3,940 chars)","verified":false,"liveness":"unknown","lastLivenessCheck":null,"agentReviews":{"count":0,"score_avg":null,"cost_usd_avg":null,"success_rate":null,"latency_p50_ms":null,"narrative_summary":null,"summary_updated_at":null},"enrichmentModel":"deterministic:skill-github:v1","enrichmentVersion":1,"enrichedAt":"2026-05-02T12:54:48.257Z","embedding":null,"createdAt":"2026-04-18T22:10:03.820Z","updatedAt":"2026-05-02T12:54:48.257Z","lastSeenAt":"2026-05-02T12:54:48.257Z","tsv":"'-1':6,200,239,470,476 '-10':126 '-2':202,332,473,478 '-3':359 '-30':375 '-45':351 '-500':133,279 '-6':367 '/../docs/installation.md':72 '/boltz-2':7 '0.5':267 '0.65':253 '0.7':262,271 '0.78':251 '0.81':255 '1':17,80,125 '100':278,348 '1000':364 '12.0':54 '12.1':55 '15':374 '1800s':100 '2':22,103,358,446 '200':131 '2024':491,493 '24gb':58 '2x':481 '3':26,124 '3.10':51 '3.11':52 '30':350 '32gb':62 '35':361 '4':34,366 '48gb':59,98 '5':40 '50':132 '500':356 '64gb':63 '70':369 '8':353 '80gb':411 'a100':410 'a100-80gb':409 'aa':462 'accuraci':321,483 'af2':33,204 'af2-multimer':203 'altern':31 'aspect':474 'b':150 'bash':82,106,384 'basic':487 'best':156 'better':488 'binder':25,380 'binder-target':379 'bio.pdb':179 'biomod':78,84 'biomolecular':10 'boltz':1,5,43,109,110,168,199,201,230,238,289,298,305,318,324,330,331,334,453,469,472,475,477 'bool':140 'campaign':341,363 'caus':420 'cd':83 'chai':300,307,326,338 'chain':441,447 'check':463 'cif':158,170,276,388 'colabfold':302,322 'comparison':197 'complex':20,39,283,297,349,357,365,373,382,427 'complex.fasta':91,113,233 'comprehens':370 'confid':263,268,272,397 'confidence.json':160 'consid':337 'convert':172 'cost':345 'count':394 'cuda':53,423 'decis':284 'default':101,119 'descript':121 'design':24 'diffus':134,213 'dir':94 'dna/rna':335 'download':454 'ensur':443 'error':402,417,419 'esmfold':311 'faa':90 'fals':434 'fast':218,480 'fasta':112,144,232,444 'faster':219,312,482 'featur':198 'file':277 'filenotfounderror':448 'filter':499 'find':385 'first':67,455 'fix':421 'format':145,153,159,171 'free':208,406 'global':264 'good':257,484 'gpu':56,96,437 'guid':71 'h':360,368 'import':180 'improv':485 'increas':398 'info':236,241,244 'input':89,393 'input-faa':88 'instal':70,105,108 'interfac':269 'interpret':418 'invalid':457 'io':190 'io.save':195 'io.set':192 'iptm':163,252,266,439 'iter':128 'kb':280 'key':116 'keyerror':438 'l':390 'l40s':60,97,344 'larg':362,429 'larger':436 'late':492 'ligand':38,304,486 'like':260 'load':237 'local':42,104 'look':259 'low':396 'match':392 'matrix':166 'maximum':320 'memori':426 'min':352 'minimum':48 'miss':450 'mktayiakqrqisfvk':148 'mmcifpars':181,184 'modal':76,81,85,346 'modal_boltz.py':87 'mode':209,407 'model':157,188,246,451 'model_0.cif':155,189 'model_0.pdb':196 'modifi':465 'msa':137,142,207,314,405,432 'msa-fre':206,404 'multim':205 'mvlspadktnvkaawg':151 'name':387 'need':27,176,313 'next':494 'non':460 'non-standard':459 'note':167,347 'oom':401 'open':9,29,221,327 'open-sourc':28 'option':79,102 'out-dir':92 'output':114,152,169,227,234,258 'pae':165 'pae.npy':164 'paramet':117,118 'parser':183 'parser.get':186 'pdb':174 'pdbio':182,191 'per':274,372 'per-complex':371 'per-residu':273 'perform':340 'pip':107 'plddt':161,254,270 'predict':3,18,35,45,95,111,115,154,231,235,242,293,386,413 'predictions/confidence.json':249 'predictions/model_0.cif':248 'predictor':12 'prefer':329 'prerequisit':46 'protein':19,37,146,149,295,296,303,309,496 'protein-ligand':36 'protein-protein':294 'protein-qc':495 'ptm':162,250,261 'python':50,177 'qc':497 'ram':61 'rang':120 'rank':501 'recommend':49 'recycl':122,127,399 'reduc':414 'releas':490 'requir':47 'residu':275,458,466 'run':66,86,229,452 'runtimeerror':422 'sampl':129,226,415 'save':245 'see':69 'sequenc':468 'server':138,143,433 'set':74 'singl':308,440 'size':342 'skill':15 'skill-boltz' 'slow':412 'slower':220 'sourc':30,222 'source-adaptyvbio' 'speed':217,317,333,479 'standard':354,461 'step':123,130,135,400,416 'structur':2,11,21,44,185,187,193,194,243,265 'success':228 'support':336,489 'target':381 'time':68,343 'timeout':99 'topic-agent-skills' 'topic-claude-code' 'topic-protein-design' 'topic-protein-engineering' 'tree':285 'troubleshoot':395 'true':139 'typic':282,339,378 'use':4,13,41,136,141,288,310,403,430,431 'valid':23,355 'valueerror':456 'verifi':383 'vram':57 'vs':471 'want':316 'wc':389 'weight':240,328,449 'yes':210,211,214,215,223,224,225,319","prices":[{"id":"106a87d7-1ee8-487b-8627-3b5f34a9b22c","listingId":"a9c24eee-34b2-4530-b0ee-1e0f206a9a59","amountUsd":"0","unit":"free","nativeCurrency":null,"nativeAmount":null,"chain":null,"payTo":null,"paymentMethod":"skill-free","isPrimary":true,"details":{"org":"adaptyvbio","category":"protein-design-skills","install_from":"skills.sh"},"createdAt":"2026-04-18T22:10:03.820Z"}],"sources":[{"listingId":"a9c24eee-34b2-4530-b0ee-1e0f206a9a59","source":"github","sourceId":"adaptyvbio/protein-design-skills/boltz","sourceUrl":"https://github.com/adaptyvbio/protein-design-skills/tree/main/skills/boltz","isPrimary":false,"firstSeenAt":"2026-04-18T22:10:03.820Z","lastSeenAt":"2026-05-02T12:54:48.257Z"}],"details":{"listingId":"a9c24eee-34b2-4530-b0ee-1e0f206a9a59","quickStartSnippet":null,"exampleRequest":null,"exampleResponse":null,"schema":null,"openapiUrl":null,"agentsTxtUrl":null,"citations":[],"useCases":[],"bestFor":[],"notFor":[],"kindDetails":{"org":"adaptyvbio","slug":"boltz","github":{"repo":"adaptyvbio/protein-design-skills","stars":126,"topics":["agent-skills","claude-code","protein-design","protein-engineering"],"license":"mit","html_url":"https://github.com/adaptyvbio/protein-design-skills","pushed_at":"2026-01-19T13:06:29Z","description":"Claude Code skills for protein design","skill_md_sha":"09fdf910b25d88c42219f934e19ba2d8f7c3819d","skill_md_path":"skills/boltz/SKILL.md","default_branch":"main","skill_tree_url":"https://github.com/adaptyvbio/protein-design-skills/tree/main/skills/boltz"},"layout":"multi","source":"github","category":"protein-design-skills","frontmatter":{"name":"boltz","license":"MIT","description":"Structure prediction using Boltz-1/Boltz-2, an open biomolecular structure predictor. Use this skill when: (1) Predicting protein complex structures, (2) Validating designed binders, (3) Need open-source alternative to AF2, (4) Predicting protein-ligand complexes, (5) Using local GPU resources.  For QC thresholds, use protein-qc. For AlphaFold2 prediction, use alphafold. For Chai prediction, use chai."},"skills_sh_url":"https://skills.sh/adaptyvbio/protein-design-skills/boltz"},"updatedAt":"2026-05-02T12:54:48.257Z"}}