Identifying motif combinations with extreme expression
Shane Chu, Chaolin Zhang (upcoming, 2026)
We introduce ATLAS, a method that identifies co-occurring motif combinations and quantifies their joint influence on predicted expression using power indices from cooperative game theory. Across eight datasets spanning alternative splicing, 5' UTRs, and promoters, we find that combinations of three to five motifs form tight clusters at the tails of expression, while single motifs do not.
Identifying interactive effects of variant combinations for protein mutagenesis
(upcoming, 2026)
A method for discovering how combinations of protein variants interact to produce functional effects beyond individual contributions.
Finding motifs using DNA images derived from sparse representations
[Bioinformatics, 2023] [Github]
A motif discovery method that identifies both gapped and cooperative binding patterns, in addition to finding the primary binding sites.