About lab
One published Lab
Build with this Lab. Run it anywhere.
Use the same immutable release locally or through managed cloud execution, with its exact version and provenance preserved.
Preparing exact examples…
Guided Boltz-2 workflow for ranking a small ligand CSV against one protein target using binder probability, affinity-like value, confidence metrics, flags, and report-ready output.
Manifest
{
"io": {
"inputs": [
{
"ui": {
"rows": 6,
"control": "textarea"
},
"name": "protein_sequence",
"label": "ABL1 Kinase Sequence",
"format": "sequence",
"maps_to": "ligand_library_loader.protein_sequence",
"required": false,
"value_type": "string",
"description": "Amino-acid sequence for the shared ABL1 kinase-domain target; maps directly to the Boltz-2 protein input for each ligand run."
},
{
"ui": {
"control": "file"
},
"file": {
"accept": [
".csv",
"text/csv"
]
},
"name": "ligand_csv",
"label": "Candidate Ligand CSV",
"format": "csv",
"maps_to": "ligand_library_loader.ligand_csv",
"required": false,
"value_type": "file",
"description": "CSV text with candidate ligand names and SMILES strings; each row is passed to Boltz-2 as a separate ligand input."
},
{
"ui": {
"control": "file"
},
"file": {
"accept": [
".a3m",
".fasta",
".fa"
]
},
"name": "msa_path",
"label": "Precomputed MSA Path",
"format": "a3m",
"maps_to": "ligand_library_loader.msa_path",
"required": false,
"value_type": "file",
"description": "Optional path to a precomputed MSA file; leave unset when the configured Boltz MSA server is used."
},
{
"name": "use_msa_server",
"label": "Use MSA Server",
"default": true,
"maps_to": "ligand_library_loader.run_options.use_msa_server",
"value_type": "boolean",
"description": "Use Boltz's MSA server instead of requiring a precomputed MSA file."
},
{
"name": "sampling_steps",
"label": "Sampling Steps",
"default": 200,
"maps_to": "ligand_library_loader.run_options.sampling_steps",
"value_type": "integer",
"description": "Number of Boltz sampling steps for structure generation."
},
{
"name": "recycling_steps",
"label": "Recycling Steps",
"default": 3,
"maps_to": "ligand_library_loader.run_options.recycling_steps",
"value_type": "integer",
"description": "Number of structure recycling steps."
},
{
"name": "diffusion_samples",
"label": "Diffusion Samples",
"default": 1,
"maps_to": "ligand_library_loader.run_options.diffusion_samples",
"value_type": "integer",
"description": "Number of Boltz diffusion samples to generate."
},
{
"name": "output_format",
"label": "Output Format",
"default": "mmcif",
"maps_to": "ligand_library_loader.run_options.output_format",
"value_type": "string",
"description": "Structure artifact format emitted by Boltz.",
"allowed_values": [
"mmcif",
"pdb"
]
},
{
"ui": {
"rows": 6,
"control": "json"
},
"name": "run_options",
"label": "Boltz Batch Run Options",
"format": "json",
"maps_to": "ligand_library_loader.run_options",
"advanced": true,
"required": false,
"value_type": "record",
"description": "Optional structured controls for the batch workflow, including source metadata, sampling settings, and interpretation scope."
}
],
"outputs": [
{
"name": "scenario_context",
"label": "Workflow Scenario Context",
"maps_to": "batch_target_context.scenario_context",
"description": "Source-backed target, ligand, use-case, provenance, and caveat context for this Boltz workflow."
},
{
"name": "assembled_boltz_request",
"label": "Assembled Boltz Request",
"maps_to": "ligand_library_loader.assembled_boltz_request",
"description": "Traceable summary of the protein, ligand, MSA, and run options prepared for Boltz."
},
{
"name": "batch_summary",
"label": "Ranked Ligand Summary",
"maps_to": "boltz_batch_ligand_ranker.batch_summary",
"description": "Ranked ligand table with binder probability, affinity-like value, confidence, status, and caveat flags."
},
{
"name": "affinity_summary",
"label": "Top Ligand Affinity-Style Summary",
"maps_to": "boltz_batch_ligand_ranker.affinity_summary",
"description": "Parsed Boltz-2 binder probability and affinity-like outputs for the top-ranked completed ligand."
},
{
"name": "confidence_summary",
"label": "Top Ligand Confidence Summary",
"maps_to": "boltz_batch_ligand_ranker.confidence_summary",
"description": "Parsed Boltz-2 confidence outputs for the top-ranked predicted complex."
},
{
"name": "structure_artifacts",
"label": "Top Ligand Structure Artifacts",
"maps_to": "boltz_batch_ligand_ranker.structure_artifacts",
"description": "File-backed predicted complex structures for the top-ranked completed ligand."
},
{
"name": "run_metadata",
"label": "Boltz Batch Run Metadata",
"maps_to": "boltz_batch_ligand_ranker.run_metadata",
"description": "Runtime status, per-ligand statuses, command metadata, logs, and caveats for the latest batch invocation."
},
{
"name": "ranking_evidence",
"label": "Conservative Ranking Evidence",
"maps_to": "ranking_interpreter.prediction_evidence",
"description": "Interpreted Boltz output evidence with request traceability and explicit scientific caveats."
}
]
},
"tags": [
"boltz-workflow",
"boltz",
"protein-ligand",
"affinity",
"structural-biology",
"guided-workflow",
"batch-ranking",
"gpu"
],
"title": "Boltz Workflow: Batch Ligand Ranking",
"models": [
{
"path": "models/context",
"alias": "batch_target_context",
"parameters": {
"scenario": {
"caveat": "Boltz-2 outputs are computational structure and affinity-style predictions for hypothesis generation; they are not experimental binding, potency, selectivity, clinical, or efficacy evidence.",
"source_pdb": "2HYY",
"ligand_role": "candidate ligand set",
"target_name": "Human ABL1 kinase domain",
"disease_area": "small ligand ranking",
"target_family": "ABL1 kinase domain",
"workflow_name": "Batch Ligand Ranking",
"workflow_context": "Guided Boltz-2 batch ligand ranking workflow using ABL1 kinase-domain examples",
"workflow_question": "How does a small ligand library rank against the shared ABL1 kinase-domain target?",
"interpretation_scope": "Learning, small-set ranking, and early hypothesis generation only",
"protein_sequence_length": 273
},
"integration_step": 0.01
}
},
{
"path": "models/input_assembler",
"alias": "ligand_library_loader",
"parameters": {
"workflow_kind": "batch",
"workflow_name": "Batch Ligand Ranking",
"integration_step": 0.01,
"default_ligand_csv": "name,smiles,source\nImatinib,CC1=C(C=C(C=C1)NC(=O)C2=CC=C(C=C2)CN3CCN(CC3)C)NC4=NC=CC(=N4)C5=CN=CC=C5,PubChem CID 5291\nDasatinib,CC1=C(C(=CC=C1)Cl)NC(=O)C2=CN=C(S2)NC3=CC(=NC(=N3)C)N4CCN(CC4)CCO,PubChem CID 3062316\nNilotinib,CC1=C(C=C(C=C1)C(=O)NC2=CC(=CC(=C2)C(F)(F)F)N3C=C(N=C3)C)NC4=NC=CC(=N4)C5=CN=CC=C5,PubChem CID 644241\n",
"default_run_options": {
"source_pdb": "2HYY",
"target_name": "Human ABL1 kinase domain",
"workflow_name": "Batch Ligand Ranking",
"ligand_examples": [
"Imatinib",
"Dasatinib",
"Nilotinib"
],
"workflow_context": "Guided Boltz-2 batch ligand ranking workflow using ABL1 kinase-domain examples",
"interpretation_scope": "Learning, small-set ranking, and early hypothesis generation only"
},
"default_protein_sequence": "VSPNYDKWEMERTDITMKHKLGGGQYGEVYEGVWKKYSLTVAVKTLKEDTMEVEEFLKEAAVMKEIKHPNLVQLLGVCTREPPFYIITEFMTYGNLLDYLRECNRQEVNAVVLLYMATQISSAMEYLEKKNFIHRDLAARNCLVGENHLVKVADFGLSRLMTGDTYTAHAGAKFPIKWTAPESLAYNKFSIKSDVWAFGVLLWEIATYGMSPYPGIDLSQVYELLEKDYRMERPEGCPEKVYELMRACWQWNPSDRPSFAEIHQAFETMFQES"
}
},
{
"path": "models/core",
"alias": "boltz_batch_ligand_ranker",
"parameters": {
"override": true,
"accelerator": "gpu",
"max_ligands": 3,
"runtime_mode": "managed",
"output_format": "mmcif",
"sampling_steps": 200,
"use_msa_server": true,
"recycling_steps": 3,
"diffusion_samples": 1
}
},
{
"path": "models/interpreter",
"alias": "ranking_interpreter",
"parameters": {
"mode": "batch",
"caveat": "Boltz-2 outputs are computational structure and affinity-style predictions for hypothesis generation; they are not experimental binding, potency, selectivity, clinical, or efficacy evidence.",
"core_alias": "boltz_batch_ligand_ranker",
"workflow_name": "Batch Ligand Ranking",
"integration_step": 0.01
}
},
{
"path": "models/visualisation",
"alias": "visualisation",
"parameters": {
"mode": "boltz_batch",
"lab_title": "Boltz Workflow: Batch Ligand Ranking",
"source_alias": "boltz_batch_ligand_ranker",
"context_alias": "batch_target_context",
"assembler_alias": "ligand_library_loader",
"integration_step": 0.01,
"interpreter_alias": "ranking_interpreter"
}
}
],
"wiring": [
{
"to": [
"ligand_library_loader.scenario_context",
"ranking_interpreter.scenario_context",
"visualisation.batch_target_context_scenario_context"
],
"from": "batch_target_context.scenario_context"
},
{
"to": [
"boltz_batch_ligand_ranker.protein_sequence"
],
"from": "ligand_library_loader.protein_sequence"
},
{
"to": [
"boltz_batch_ligand_ranker.ligand_csv"
],
"from": "ligand_library_loader.ligand_csv"
},
{
"to": [
"boltz_batch_ligand_ranker.msa_path"
],
"from": "ligand_library_loader.msa_path"
},
{
"to": [
"boltz_batch_ligand_ranker.run_options"
],
"from": "ligand_library_loader.run_options"
},
{
"to": [
"ranking_interpreter.assembled_boltz_request",
"visualisation.ligand_library_loader_assembled_boltz_request"
],
"from": "ligand_library_loader.assembled_boltz_request"
},
{
"to": [
"visualisation.boltz_batch_ligand_ranker_batch_summary",
"ranking_interpreter.boltz_batch_ligand_ranker_batch_summary"
],
"from": "boltz_batch_ligand_ranker.batch_summary"
},
{
"to": [
"visualisation.boltz_batch_ligand_ranker_affinity_summary",
"ranking_interpreter.boltz_batch_ligand_ranker_affinity_summary"
],
"from": "boltz_batch_ligand_ranker.affinity_summary"
},
{
"to": [
"visualisation.boltz_batch_ligand_ranker_confidence_summary",
"ranking_interpreter.boltz_batch_ligand_ranker_confidence_summary"
],
"from": "boltz_batch_ligand_ranker.confidence_summary"
},
{
"to": [
"visualisation.boltz_batch_ligand_ranker_structure_artifacts",
"ranking_interpreter.boltz_batch_ligand_ranker_structure_artifacts"
],
"from": "boltz_batch_ligand_ranker.structure_artifacts"
},
{
"to": [
"visualisation.boltz_batch_ligand_ranker_run_metadata",
"ranking_interpreter.boltz_batch_ligand_ranker_run_metadata"
],
"from": "boltz_batch_ligand_ranker.run_metadata"
},
{
"to": [
"visualisation.ranking_interpreter_prediction_evidence"
],
"from": "ranking_interpreter.prediction_evidence"
}
],
"package": "boltz-batch-ligand-ranking-workflow",
"runtime": {
"duration": 0.01,
"settle_steps": 1,
"initial_inputs": {},
"python_version": "3.10",
"communication_step": 0.01
},
"version": "1.0.0",
"description": "Guided Boltz-2 workflow for ranking a small ligand CSV against one protein target using binder probability, affinity-like value, confidence metrics, flags, and report-ready output.",
"schema_version": "2.0"
}Runtime
Duration0.01
Comms Step0.01
Settle Steps1
Runs
Total0
Completed0
Failed0
Metadata
Packageboltz-batch-ligand-ranking-workflow
Created2026-05-23
Updated2026-06-13
boltz-workflowboltzprotein-ligandaffinitystructural-biologyguided-workflowbatch-rankinggpucontextprovenanceinput-assemblyinterpretationdockingvisualisationother
No public runs yet.