Mona Singh

The paper introduces Domino, a deep learning framework that models sequence-to-function relationships to uncover the context-specific grammar of Drosophila chromatin insulation.

This paper introduces ARIES, a method using protein language model embeddings to construct multiple sequence alignments, achieving higher accuracy than existing methods, especially in low-identity regimes.

Dyscovr identifies driver-associated transcriptional changes and prioritizes genes for therapeutic targeting in cancer.

The paper benchmarks a high-efficiency prime editing platform for multiplexed dropout screening of genetic variants, identifying negative growth phenotypes for nonsense mutations and splice-disrupting synonymous mutations.

This paper develops a machine learning framework to assess and reduce data bias in MHC binding predictions across racial and ethnic groups.

This paper introduces a novel multiple sequence alignment method using protein language models to improve alignment accuracy in the twilight zone of sequence similarity.

This paper introduces rCLAMPS, a method to infer protein-DNA recognition codes from DNA-binding specificities, enabling prediction of TF specificities and mutation effects.

The paper introduces NetID, a global network optimization method for annotating untargeted metabolomics data, identifying novel metabolites in yeast and mouse datasets.

The paper introduces PertInInt, a method that identifies cancer genes by analyzing somatic mutations in interaction sites across 10,037 tumor samples, revealing disrupted interactions as a common cancer mechanism.

The paper introduces a method to identify ligand-binding positions within protein domains using co-complex structures, resulting in a resource called InteracDome that aids in understanding disease-related mutations.

The paper introduces a probabilistic method for domain prediction that considers directional domain context, improving prediction accuracy by 15% over existing methods.