Contemporary Genomic Foundation Models (GFMs) rely on a DNA-as-a-string paradigm that employs masked token prediction objectives for pretraining. However, this abstraction does not explicitly model the biochemical, structural, and physical properties essential to biological function. Many molecular properties can be es...
Jeremy Levy, Ariel Larey, Yury Nahshan et al.· 0 citations
A reference-aligned diploid encoding for single-nucleotide variants (SNVs) and short insertions and deletions (indels) and unphased and phase-retaining tokenizers that accept phased genotypes and convert them to single-sequence diploid representation are introduced.
Guy Leib, T. Zinger, Dan Ofer et al.· bioRxiv· 1 citation
Protein language models are trained primarily with masked language modeling (MLM), which predicts masked amino-acid identities, but joint-embedding predictive architectures (JEPA) instead predict latent representations, but have not been applied to proteins.
Dan Ofer, Dafna Shahaf, M. Linial· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.