Masrur Sobhan

EMoMiS: A Pipeline for Epitope-based Molecular Mimicry Search in Protein Structures with Applications to SARS-CoV-2

EMoMiS identifies molecular mimicry between SARS-CoV-2 proteins and viral/human proteins, offering insights into cross-protection and side effects.

PathPCNet: Pathway Principal Component-Based Interpretable Framework for Drug Sensitivity Prediction

PathPCNet predicts drug sensitivity using multi-omics data and biological pathways, achieving high correlation and uncovering key gene-drug interactions.

PathX-CNN: An Enhanced Explainable Convolutional Neural Network for Survival Prediction and Pathway Analysis in Glioblastoma

PathX-CNN improves survival prediction and pathway analysis in glioblastoma by integrating multi-omics data with sub-cohort-specific principal components and SHAP explanations.

Interpreting Lung Cancer Health Disparity at Transcriptome Level

This study identifies biomarker genes and pathways contributing to lung cancer health disparities using a classification framework with SHAP to interpret disparities between racial and sex groups.

MOGAT: An Improved Multi-Omics Integration Framework Using Graph Attention Networks

MOGAT improves multi-omics integration for cancer subtype prediction by addressing limitations in existing graph neural network approaches.

Epitopedia: identifying molecular mimicry between pathogens and known immune epitopes

Epitopedia identifies regions of molecular mimicry between pathogens and immune epitopes by analyzing structural similarity and sequence identity.

Quantifying Intratumor Heterogeneity by Key Genes Selected Using Concrete Autoencoder

The paper develops a deep learning framework using Concrete Autoencoder to identify key genes for quantifying intratumor heterogeneity, showing that a subset of genes suffices for accurate ITH scoring.

Potential autoimmunity resulting from molecular mimicry between SARS-CoV-2 Spike and human proteins

The paper identifies molecular mimicry between SARS-CoV-2 Spike and human proteins that may contribute to autoimmune responses in COVID-19.

Multi-run Concrete Autoencoder to Identify Prognostic lncRNAs for 12 Cancers

This paper introduces a multi-run Concrete Autoencoder (mrCAE) to identify prognostic lncRNAs across 12 cancers, achieving 95% accuracy in cancer classification and identifying 76 prognostic lncRNAs.