Yevgeni Nogin

OM2Seq improves OGM data analysis by using a Transformer-encoder to rapidly and accurately map DNA fragments to a reference genome, outperforming existing methods in speed and accuracy.

This paper develops an information theory-based model to design optimal labeling patterns for optical genome mapping, improving accuracy by up to 10-fold in human and bacterial DNA.

DeepOM uses deep learning to improve alignment of single-molecule optical genome mapping images to genomic references, showing better performance for shorter DNA molecules.