Vipul Periwal

The paper examines how islet size, endocrine cell composition, and spatial location are interconnected in human pancreata using single-cell analysis of non-diabetic donors.

The paper introduces MESIC, a method that maps gene expression data into an interpretable semantic space using gene annotations, enabling biological interpretation of clusters and outliers.

This paper introduces DynML, a dynamical systems framework for modeling biological time series, showing that reservoir topological entropy correlates with predictive power.

The paper develops a probabilistic model to forecast HIV protease drug resistance by simulating evolutionary paths and inferring epistatic interactions from clinical data.

The paper develops a hybrid deep learning-mechanistic model to infer gene regulatory dynamics from a spatiotemporal single-cell atlas of Drosophila embryos, enabling interpretable predictions of gene interactions.

This paper introduces SensX, an XAI framework that improves explanation accuracy and speed for deep learning models, validated on vision transformers and single-cell RNA-seq data.

This study develops a personalized digital twin model for liver recovery using gene expression data from LDLT donors to predict individual recovery trajectories.

This paper develops a deep learning framework for parameter inference in physiological models using simulated data and clinical glucose tolerance testing data.

This paper introduces a parameter-free statistical physics-based method for protein structure prediction, achieving better performance than existing methods for 76% of analyzed proteins.

The paper computes genome-wide covariation in SARS-CoV-2 using Expectation Reflection and analyzes relationships within and between encoded proteins across clades.

The paper infers mechanistic gene networks from single-cell data and shows they outperform neural networks in predicting gene dynamics and knockout effects.