
Explainable QSAR & SHAP
Exploring how molecular descriptors influence QSAR model predictions.


Research interests in biomedical AI and machine learning.
I am a bioinformatics researcher interested in biological data analysis and computational methods for biomedical research. My work and interests include genomics, transcriptomics, biomedical AI, and computational drug discovery.
An MS research framework exploring brain tumor histology and gene-expression data with deep learning.
Select an area to learn what it investigates:
Explores whether combining image and gene-expression information supports classification compared with using either data type alone.
Diagnostic H&E slides tiled and feature-extracted.
Normalized transcriptomic count matrices.
Aligned multimodal classification output.
Paired whole-slide diagnostic images and bulk RNA-seq count matrices from The Cancer Genome Atlas.
GBM-only cases used for external transfer testing; this is not an external two-class validation.
NeuroBridge is an academic research prototype. It is not an FDA/CE-cleared diagnostic tool and is not intended for clinical patient decision-making. No synthetic WSIs are generated.
My interests cover computational approaches to biological research, from molecular data analysis to machine learning.
Computational analysis of biological sequences, gene-expression profiles, and molecular data.
Exploring machine-learning methods for biological and biomedical datasets.
Analyzing genomic and gene-expression data to study biological variation.
Studying protein structures and molecular interactions through modeling and simulation.

Exploring how molecular descriptors influence QSAR model predictions.

Comparing molecular properties and predicted ADME or toxicity endpoints.

Comparing microbial communities through diversity measures and ordination.

Exploring possible effects of sequence variants on protein structure.

Examining predicted ligand binding poses and protein interactions.

Building comparative protein models and reviewing structural quality.
Select an expertise domain below to inspect core methodologies, specialized scientific software, and corresponding verified investigations.
Selected tools and methods used or explored in academic and computational work.
Completed the NeuroBridge thesis on histopathology and gene-expression analysis.
Contributed to computational research involving histopathology images, gene-expression data, and machine-learning analysis during the NeuroBridge project.
Studied biochemistry and investigated HIV-1 integrase inhibitors using QSAR and docking.
SINES, NUST
An MS thesis on multimodal analysis of CNS tumor histology and gene-expression data.
Department of Biochemistry, AWKUM
A bachelor's thesis studying potential HIV-1 integrase inhibitors with QSAR and molecular docking.
Manuscript submission (2026)
A structural bioinformatics study of selected CD28 variants and their possible effects.
I welcome discussions about multimodal biomedical AI, computational biology, and research collaborations.