Statistics
PM3-related evidence sparsity
PM3-related evidence was sparse relative to the full variant–publication search space. Among 103,889 canonical records, 3,842 (3.70%) were classified as PM3-related evidence, including 223 phase-confirmed records and 3,619 phase-unconfirmed biallelic evidence records. These records involved 2,759 index variants across 701 genes, indicating that PM3-relevant observations are distributed across a broad gene space but remain uncommon at the level of individual variant–publication pairs.
This sparsity highlights a central challenge for PM3-oriented interpretation: clinically relevant genotype evidence is scattered across a large literature corpus, while most candidate variant–publication associations do not provide usable PM3 evidence. A dedicated PM3 database is therefore valuable not because PM3 evidence is abundant, but because it is rare, fragmented, and difficult to identify efficiently through publication-level search alone.
Allelic configuration context distribution
Reviewed source distribution
Publication year distribution
Data quality summary
Release-level quality assessment includes benchmark evaluation, manual audit and database consistency checks. Detailed methods and results belong in the manuscript and supplementary materials; this page summarizes database scale, evidence distribution and source provenance for portal users.