All Competitors

A short-context masked DNA language model trained on curated regulatory sequences with a motif-discovery regularizer for zero-shot TF motif recovery and variant effect prediction.

BERT-based model pretrained on 15-state ROADMAP chromatin annotations across 127 human cell types to uncover chromatin-state motifs and predict gene expression.

Motif-oriented DNA pre-training framework using an ELECTRA-style generator-discriminator architecture to learn biologically informed genomic representations.