August 5, 2025Open Access

Benchmarking DNA Foundation Models for zero-shot variant effect prediction: the role of context, training, and architecture

View Full Paper

Ask AI

Bookmark

Authors

Discussion

Benchmarking dna models improves variant predictions using zero-shot scoring, highlighting model architecture's role.

Large multi-species dna models outperform human-only models in variant effect prediction.
Models effectively capture local and extended contextual effects using zero-shot scoring on snvs.
The study uses data from Clin-Var to assess the functional consequences of pathogenic, benign, and uncertain variants.
Innovations in transformer architecture may enhance modeling of long-range dependencies in genomic sequences.

Alfisi et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark

August 5, 2025Open Access

View Full Paper

Ask AI

Bookmark

Authors

Benchmarking dna models improves variant predictions using zero-shot scoring, highlighting model architecture's role.

Large multi-species dna models outperform human-only models in variant effect prediction.
Models effectively capture local and extended contextual effects using zero-shot scoring on snvs.
The study uses data from Clin-Var to assess the functional consequences of pathogenic, benign, and uncertain variants.
Innovations in transformer architecture may enhance modeling of long-range dependencies in genomic sequences.

Alfisi et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark