{ "blueprint_schema": "nmd-vcell-competitive-upgrade-blueprint/1.0", "resource_release": "v1.0.0-database-resource", "evidence_freeze": "2026-07-25", "interface_build": "EA-20260730-42", "source_basis": "User-attached competitive benchmark report dated 2026-07-31; external competitor facts are treated as design references, not inherited evidence.", "generated_by": "scripts/apply-evidence-atlas-product-overlay.mjs", "positioning": "NMD-VCell should become a disease-specific virtual-cell evidence, prediction-registration and prospective-validation platform for neuromuscular disease, not a generic foundation-model leaderboard.", "lessons": [ { "source_pattern": "Arc STATE and Virtual Cell Challenge", "learn": "A model claim must be bound to task, split, hidden labels, frozen submission and permanent simple baselines.", "borrow": "Prediction-freeze, hidden-test and baseline-retention rules.", "apply_to_nmd_vcell": "Create a DMD Prospective Benchmark where Study Card, Prediction Card and Outcome object are separate immutable records.", "current_status": "P0_INFRASTRUCTURE_READY_P1_LOCKED", "acceptance_gate": "No DMD prediction label is shown until a timestamped prediction exists before outcome access and simple baselines run on the same split." }, { "source_pattern": "CZI Virtual Cells Platform", "learn": "Data, model, benchmark and inference should use one ModelRun contract across web, CLI and local execution.", "borrow": "Model adapter, run log, dataset digest, split manifest and reusable benchmark package.", "apply_to_nmd_vcell": "Standardize zero, train-mean, nearest-neighbour, ridge and one future population model behind a shared ModelRun object.", "current_status": "MODEL_CARD_ONLY_FOR_ADVANCED_DMD_MODELS", "acceptance_gate": "Every run records model id, commit, container digest, data SHA, split id, seed, status, metrics and failure reason." }, { "source_pattern": "CELLxGENE and Vitessce-style browsers", "learn": "A Cell Browser must let users operate real cell objects, not only read concept cards.", "borrow": "Dataset, donor, disease, state, batch, expression, DE, cell-proportion and spatial-neighbour filters.", "apply_to_nmd_vcell": "Upgrade Cell Browser v1 around real DMD scRNA/snRNA/spatial data with donor-level pseudobulk as the default statistical unit.", "current_status": "CONTEXT_MAP_NOT_CELL_LEVEL_BROWSER", "acceptance_gate": "At least two DMD datasets can be filtered by donor and disease, with expression, cell count, pseudobulk DE and exportable selections." }, { "source_pattern": "Open Targets and DepMap", "learn": "Long-lived platforms are entity graphs plus stable programmatic file indices, not only pages.", "borrow": "Entity relation graph, API version metadata, file index, historical releases and provenance by source.", "apply_to_nmd_vcell": "Expose Gene, Dataset, Comparison, Study, ModelRun, Prediction and Outcome as linked objects under one release context.", "current_status": "PARTIAL_OBJECT_GRAPH_RELEASED", "acceptance_gate": "Each page answers: what object is studied, what evidence supports it, and which next object would change the conclusion." }, { "source_pattern": "VCell mechanism simulation platform", "learn": "Mechanistic models separate model definition, parameters, solver, simulation run and result object.", "borrow": "Solver/run/result separation and explicit parameter documentation.", "apply_to_nmd_vcell": "Keep Computability Gate, ModelRun and future mechanistic simulation objects distinct so the name virtual cell does not imply unavailable numerical simulation.", "current_status": "COMPUTABILITY_GATE_ONLY", "acceptance_gate": "UI never presents evidence checks as simulation output; future solver objects must cite equations, parameters, run logs and result files." }, { "source_pattern": "CLUE/CMap service continuity lesson", "learn": "A valuable scientific portal can lose utility if data are only reachable through an active website.", "borrow": "Portable archive, checksum ledger, mirror plan, historical manifest and exit strategy.", "apply_to_nmd_vcell": "Keep RO-Crate, release manifest, file registry, checksums and deprecated archive policy as first-class release objects.", "current_status": "PORTABLE_RELEASE_OBJECTS_PRESENT_FORMAL_DOI_PENDING", "acceptance_gate": "A release can be reconstructed from manifest, checksums, schemas and archived files without relying on the interactive site." } ], "implementation_axes": [ { "axis": "Release engineering", "current_state": "Health, route manifest and overlay identity are generated and checked.", "upgrade_action": "Keep manifest, downloads, overlay, health and route QA as a single generated release context.", "status": "IMPLEMENTED_INTERNAL_GATE", "evidence_objects": [ "api/health", "api/v1.1/release_manifest.json", "api/v1.1/overlay_archive_identity.json", "api/v1.1/route_manifest.json" ] }, { "axis": "Deep-link research objects", "current_state": "Planner can restore gene state and object trace server-side.", "upgrade_action": "Extend parent inheritance from Comparison to Study Card, Prediction and Outcome object URLs.", "status": "PARTIAL_IMPLEMENTED", "evidence_objects": [ "/resource/planner?gene=ZNF133", "api/v1.1/study-cards/ZNF133.json" ] }, { "axis": "Real Cell Browser", "current_state": "Cell Browser is a disease ecology and computability map.", "upgrade_action": "Connect real DMD single-cell, single-nucleus and spatial datasets with donor-level filters and pseudobulk DE.", "status": "PLANNED_REQUIRES_DATA_IMPORT", "evidence_objects": [ "api/v1.1/dataset_registry.json" ] }, { "axis": "Executable ModelRun system", "current_state": "Model Cards and benchmark contracts exist; calibrated DMD model count is zero.", "upgrade_action": "Run permanent baselines and future population model under one adapter and immutable ModelRun schema.", "status": "PLANNED_REQUIRES_RUNNER", "evidence_objects": [ "api/v1.1/model_registry.json", "api/v1.1/benchmark_suite.json" ] }, { "axis": "Prospective validation loop", "current_state": "Study Cards are drafts; predictions and measured outcomes are zero.", "upgrade_action": "Create append-only Study -> Prediction -> Experiment -> Outcome chain with null, toxic, failed-QC and abstained outcomes preserved.", "status": "PLANNED_REQUIRES_EXPERIMENT_AND_WRITE_PATH", "evidence_objects": [ "api/v1.1/study_card_manifest.json", "api/v1.1/prediction_registry.json" ] } ], "research_object_v2_contract": { "envelope_fields": [ "object_id", "object_type", "schema_version", "lifecycle_status", "release_context", "identity.parents", "biological_context", "perturbation", "endpoint", "evidence_state", "provenance", "files", "validation" ], "dataset_required_fields": [ "accession", "permanent_identifier", "license", "species", "disease", "donor_count", "cell_count", "modality", "perturbation_type", "endpoint", "raw_sha256", "processed_sha256", "benchmark_roles", "permitted_model_uses", "prohibited_uses" ], "model_run_required_fields": [ "model_id", "model_version", "source_commit", "checkpoint_sha256", "container_digest", "data_release", "feature_schema", "split_id", "random_seed", "status", "failure_reason", "predicted_object_ids", "metric_object_ids", "stdout_log_sha256", "stderr_log_sha256" ] }, "milestones": [ { "horizon": "0-30 days", "gate": "Release context, health, route contract, deep-link SSR and missing-data/reproducibility inventory.", "nmd_status": "MOSTLY_IMPLEMENTED_IN_V93_INTERNAL_GATE" }, { "horizon": "1-3 months", "gate": "DMD Cell Browser v1, Dataset Registry v2, ModelAdapter baselines and Study Card state machine.", "nmd_status": "NEXT_SOFTWARE_PHASE" }, { "horizon": "3-6 months", "gate": "Minimum Perturbome freeze, hidden prospective roles, Prediction Card and Outcome alpha.", "nmd_status": "REQUIRES_GOVERNED_STUDY_DESIGN" }, { "horizon": "6-12 months", "gate": "At least one independent DMD Study -> Prediction -> Outcome chain with failures preserved.", "nmd_status": "REQUIRES_EXPERIMENTAL_OUTCOME" } ], "claims_not_made": [ "Arc-scale data parity", "CZI-scale model inference platform parity", "CELLxGENE-level cell browser already implemented", "validated DMD perturbation prediction", "clinical decision support", "therapeutic target ranking" ], "claim_boundary": "v1.0 is a database and evidence-governance release; it is not a validated disease-prediction or clinical decision-support release." }