Shahin Mohammadi

This study evaluates foundation models for predicting cellular responses to perturbations, finding that some pretrained embeddings improve predictions while others do not.

This paper benchmarks foundation model representations for predicting cellular perturbation responses, identifying effective representation learning strategies.

The paper creates a single-cell atlas of the human primary motor cortex in ALS and FTLD, identifying transcriptional changes and molecular similarities between Betz cells and VENs.

The paper uses single-cell transcriptomics to analyze the hippocampus and entorhinal cortex in Alzheimer's disease, identifying stage-specific molecular changes across cell types.

The paper deconvolves 3,387 post-mortem brain samples using SPLITR to identify cell-type-specific eQTLs and genetic effects in mental disorders.

ACTIONet is a multiresolution framework that uses archetypal analysis and network theory to characterize single-cell state landscapes across tissues and organisms.

The paper introduces SCINET, a framework to reconstruct cell-type-specific interactomes from single-cell data, revealing disease-specific gene networks and cell-type convergence in brain disorders.

The paper introduces ACTION, a method to infer cell functional identity from transcriptomic data, identifies novel melanoma subtypes with distinct survival and therapeutic responses, and provides biomarkers and regulatory networks.

The paper introduces DECODE, an algorithm that identifies gene presence/absence patterns in single-cell RNA-seq data and distinguishes biological from technical zeros, improving functional inference and imputation.