field	required	value_type	allowed_values	definition	blocking_gate
record_id	yes	string	unique stable identifier	Stable identifier for this evidence record.	schema
gene	yes	string	HGNC-style symbol	Gene evaluated by the external evidence.	schema
event_date	yes	date	YYYY-MM-DD	Date on which the evidence state became available.	schema
study_id	yes	string	stable project or study identifier	Identifier that links all records from the same external study.	schema
record_status	yes	enum	candidate_import	Declares that the row is proposed evidence and not yet canonical.	schema
evidence_scope	yes	string	free text	Short scope statement for the external evidence.	schema
biological_context	yes	string	free text	Disease, tissue, cell state and relevant treatment context.	context
species	yes	string	free text	Species in which the endpoint was measured.	context
model_system	yes	string	free text	Model system, cohort or assay platform.	context
perturbation_type	yes	string	free text	Genetic, pharmacological or observational perturbation.	context
comparator	yes	string	free text	Exact comparison used for the effect estimate.	context
endpoint_name	yes	string	free text	Primary endpoint represented by the imported row.	context
endpoint_unit	yes	string	free text	Scale or unit of the endpoint.	statistics
endpoint_direction	yes	enum	higher_is_better;lower_is_better;two_sided	Direction used to interpret the endpoint.	statistics
result_class	yes	enum	no_material_signal;supportive_signal;mixed;inconclusive	Submitter classification; the preflight independently checks whether it is justified.	statistics
effect_measure	yes	string	free text	Name of the effect measure, such as mean difference or log2 fold change.	statistics
effect_estimate	yes	number	finite decimal	Observed effect estimate on the declared endpoint scale.	statistics
ci_level	yes	number	0 to 1	Confidence level as a proportion, for example 0.95.	statistics
ci_lower	yes	number	finite decimal	Lower confidence-interval bound on the effect scale.	statistics
ci_upper	yes	number	finite decimal	Upper confidence-interval bound on the effect scale.	statistics
standard_error	no	number	finite decimal	Standard error when available.	statistics
p_value	no	number	0 to 1	Exact unadjusted P value; blank if not calculated.	statistics
q_value	no	number	0 to 1	Exact adjusted P or q value when multiplicity applies.	statistics
n_hypotheses_tested	yes	integer	1 or greater	Number of hypotheses in the declared multiplicity family.	statistics
multiple_testing_method	conditional	string	method name	Required when more than one hypothesis was tested.	statistics
n_units	yes	integer	1 or greater	Number of independent biological or observational units.	statistics
n_independent_replicates	yes	integer	1 or greater	Number of independent experimental or cohort replicates.	statistics
null_evidence_framework	yes	enum	equivalence_test;confidence_interval_against_predeclared_margin;noninferiority_test;absence_not_tested	Framework used to support or decline a no-material-signal statement.	statistics
negligible_effect_lower	conditional	number	finite decimal	Predeclared lower bound of the negligible-effect region; required for no_material_signal.	statistics
negligible_effect_upper	conditional	number	finite decimal	Predeclared upper bound of the negligible-effect region; required for no_material_signal.	statistics
equivalence_p_value	no	number	0 to 1	Exact equivalence or noninferiority P value when that framework was used.	statistics
analysis_design	yes	string	free text	Analysis model, covariates, unit of analysis and uncertainty procedure.	reproducibility
determinism_class	yes	enum	deterministic;stochastic_seeded;environment_sensitive	Reproducibility class of the analysis.	reproducibility
context_match_to_nmd_vcell	yes	enum	direct_same_endpoint_and_context;partial_context_match;cross_context_only;unknown	Declared match between the imported evidence and the NMD-VCell candidate claim.	context
independence_from_nmd_vcell	yes	enum	independent_external_project;partially_overlapping_inputs;same_project_reanalysis;unknown	Whether evidence is statistically and procedurally independent of NMD-VCell.	context
data_reference	yes	string	path, accession or immutable URL	Reference that resolves the analyzed data object.	provenance
data_sha256	yes	sha256	64 lowercase hexadecimal characters	SHA-256 of the exact analyzed data object or frozen manifest.	provenance
code_reference	yes	string	path, repository or archive reference	Reference that resolves the analysis code.	provenance
code_version	yes	string	commit or immutable version	Exact code version used for the result.	provenance
code_sha256	yes	sha256	64 lowercase hexadecimal characters	SHA-256 of the code archive or frozen code manifest.	provenance
environment_reference	yes	string	lockfile, container digest or environment manifest	Reference to the computational environment.	reproducibility
source_reference	yes	string	accession, DOI, report or immutable record	Traceable source for the evidence and its interpretation.	provenance
source_reference_type	yes	enum	accession;doi;repository_release;internal_frozen_report;other_traceable	Kind of traceable source reference.	provenance
author_attestation	yes	boolean	true;false	Attestation that the submitted row represents the stated source without selective omission.	provenance
notes	no	string	free text	Limitations or other context; never used to override a failed gate.	none