{"stable_id":"027158736e730ddc","slug":"alphafold-protein-structure-database","company":"Google DeepMind and EMBL-EBI","workflow_name":"AlphaFold Protein Structure Database precomputed structure access","function_code":"engineering","pattern_codes":["continuous_decisioning"],"changed_assumption":"Researchers can use precomputed structure predictions, then decide what experimental validation is required.","evidence_strength":"verified","publication_tier":"showcase","freshness":"current","reviewed_at":"2026-08-23","updated_at":"2026-08-22","source_quality_summary":"1 peer reviewed, 1 primary; publication outcomes are verified.","caveat_summary":"Coverage is not equivalent to experimental accuracy or research impact; predictions have confidence limits and do not replace validation for many uses.","freshness_caveat":null,"workflow_summary":{"before":[{"actor":"Structural biologist","action":"Search the much smaller experimental structure corpus or initiate a bespoke structure-determination project.","actor_type":"control"},{"actor":"Research team","action":"Wait for a structure before pursuing structure-informed hypotheses.","actor_type":"human"}],"hinge":"Researchers can use precomputed structure predictions, then decide what experimental validation is required.","after":[{"actor":"AlphaFold DB","action":"Serves a precomputed predicted structure and confidence metrics for a queried protein sequence.","actor_type":"system"},{"actor":"Researcher","action":"Uses predictions to form hypotheses, design experiments, or solve structures, validating consequential uses.","actor_type":"system"}],"decision_mode":"moved","decision_marker":"Selection moves from a fixed rule to the model"},"before":[{"order":1,"actor":"Structural biologist","action":"Search the much smaller experimental structure corpus or initiate a bespoke structure-determination project.","handoff_to":"Experiment or computation","control":"Experimental feasibility"},{"order":2,"actor":"Research team","action":"Wait for a structure before pursuing structure-informed hypotheses.","handoff_to":"Downstream analysis","control":"Scientific judgment"}],"after":[{"order":1,"actor":"AlphaFold DB","action":"Serves a precomputed predicted structure and confidence metrics for a queried protein sequence.","handoff_to":"Researcher","control":"Prediction confidence and coverage"},{"order":2,"actor":"Researcher","action":"Uses predictions to form hypotheses, design experiments, or solve structures, validating consequential uses.","handoff_to":"Downstream research","control":"Human scientific validation"}],"decision_rights":"AlphaFold supplies a prediction; scientists decide whether confidence is sufficient and what experimental validation is required.","exception_path":"Low-confidence regions, multimers, nonstandard sequences, and consequential claims require alternative methods or experiments.","removed_work":["Bespoke first-pass prediction for proteins already in the database","Some experimental work whose only purpose was an initial structural hypothesis"],"outcomes":[{"metric":"Protein sequences with openly accessible predicted structures","baseline":"Approximately 200,000 PDB structures representing about 60,000 unique sequences","result":"More than 214 million predicted structures, covering nearly the full UniProt database","period":"Database expansion from 2021 through 2024","scale":"Global open database used by researchers in 190 countries","attribution_caveat":"Coverage is not equivalent to experimental accuracy or research impact; predictions have confidence limits and do not replace validation for many uses.","evidence_label":"verified"}],"executive_lesson":"The transformation is infrastructural: a once-scarce intermediate artifact became available on demand, shifting scientists from acquisition toward interpretation and validation.","anti_pattern":"Treating a predicted structure as experimentally confirmed or counting database entries as discoveries.","questions_for_leaders":["Which intermediate artifact could become shared AI infrastructure?","How will users see uncertainty at the point of use?"],"collections":["embodied-work"],"bundle_version":"1.0.0","bundle_fingerprint":"sha256:c23c6cc2b88153d008ea8fda928f632ce0011fc2d4c5036672a16e5d895bab93","canonical_url":"https://brianletort.ai/transformations/alphafold-protein-structure-database"}