Barak Raveh

This paper introduces iMSMs to model biomolecular interactions, revealing kinetic states and transport pathways in nucleocytoplasmic transport.

CLIPepPI predicts domain-peptide specificity using contrastive learning, achieving competitive performance across benchmarks and enabling proteome-wide analysis.

DeepDynamics uses bulk RNA-seq and sc/snRNA-seq to infer cell-subtype and clinicopathological dynamics in Alzheimer's disease, identifying risk mediators in APOE4 carriers.

This paper introduces Annotatability, a method that uses deep network training dynamics to identify and correct ambiguous cell annotations in single-cell and spatial omics data.

This paper introduces a Bayesian method to optimize metamodels for complex biological systems, showing improved accuracy in modeling beta-cell function in T2D.

The paper develops a Bayesian metamodel of pancreatic islet architecture and function by integrating imaging and physiological data with probabilistic graphical models and automation tools.