Northwestern University

Species-aware DNA embedding model built on DNABERT-2, using contrastive learning to cluster and differentiate genomic sequences by species without labeled data.

Multi-species genomic foundation model replacing k-mer tokenization with BPE, achieving state-of-the-art performance with 21x fewer parameters than prior leading models.

BERT-based pre-trained model for DNA sequences using k-mer tokenization. Achieves state-of-the-art performance on promoter, splice site, and transcription factor binding prediction.