Evidence item

The primary Nature article reports state-of-the-art performance across coding and non-coding human variant-effect prediction tasks using vertebrate, mammal, and primate alignments, with task-dependent timescale advantages and separate analyses for five model organisms. DIRECT_TEXT

Authenticity UNVERIFIED · extraction UNVERIFIED · independence group none (unreviewed) · created seq 25655

Source: Predicting genome-wide functional constraints with GPN-Star · locator {} · excerpt hash sha256:705ec6f8f669981fc6b45d9d865b4804208d04b87acbf157881a6584f5c13ba6

Exact excerpt

Trained on alignments spanning vertebrate, mammal and primate evolutionary timescales, GPN-Star achieves state-of-the-art performance across a wide range of variant effect prediction tasks in both coding and non-coding regions of the human genome. Analyses across timescales show task-dependent advantages of modelling more recent versus deeper evolution. Extending beyond humans, we train GPN-Star for five model organisms—Mus musculus, Gallus gallus, Drosophila melanogaster, Caenorhabditis elegans and Arabidopsis thaliana—demonstrating the robustness and generalizability of the framework.

Linked claims

Claim	Direction	Relevance	Steps	Counted
029e4166	SUPPORT	DIRECT	0	yes

Audits

None.