I am Hazrat Maghaz, a bioinformatics researcher with an MS in Bioinformatics from the National University of Sciences & Technology (NUST SINES), Pakistan, and a BS in Biochemistry from Abdul Wali Khan University Mardan.
My research interests include computational biology, biomedical AI, genomics, transcriptomics, and computational drug discovery. My master's research focused on integrating histopathology images and gene-expression data for CNS tumor analysis. I have also worked on computational studies involving RNA-seq analysis, QSAR modeling, protein structures, molecular docking, and molecular dynamics.
I am particularly interested in reproducible computational workflows, careful model evaluation, and biological interpretation of analytical results.
Research focus: Connecting biological data, computational methods, and biomedical questions.
| Area | Focus |
|---|---|
| Bioinformatics & Computational Biology | Biological sequence analysis, gene-expression data, molecular datasets, and reproducible computational workflows. |
| Biomedical AI & Machine Learning | Machine learning, deep learning, multimodal representation learning, and evaluation of models for biomedical data. |
| Genomics & Transcriptomics | RNA-seq analysis, differential gene expression, pathway analysis, and interpretation of molecular variation. |
| Structural Bioinformatics & Drug Discovery | Protein structure analysis, QSAR modeling, molecular docking, molecular dynamics, and pharmacophore methods. |
NeuroBridge: A Bidirectional Multimodal Deep Learning Framework for CNS Tumor Analysis Using Histopathology and Gene-Expression Profiles
MS Thesis | NUST SINES | Approved September 2026
NeuroBridge is a research framework that examines how histopathology images and gene-expression profiles can be analyzed together using deep learning. Its main classification task concerns GBM-like versus LGG-like tumor patterns, rather than all CNS tumor types.
The study used paired data from TCGA-GBM and TCGA-LGG for development and internal evaluation. CPTAC-GBM, a GBM-only cohort, was used for external transfer testing; it was not an external two-class validation cohort.
The framework also includes:
- Path2Omics: Prediction of selected molecular targets from image-derived features, not reconstruction of full RNA profiles.
- Gene2Morph: Retrieval of real histology reference patches guided by gene-expression data, not generation of synthetic whole-slide images.
- Missing-modality evaluation: Assessment of model behavior when histopathology or molecular information is unavailable.
Research overview: NeuroBridge on my academic website
NeuroBridge is an academic research prototype and is not intended for clinical diagnosis.
These are tools and methods used or explored in academic and computational projects, not proficiency ratings.
| Project | Methods | Description |
|---|---|---|
| ** Crohn's Disease RNA-seq Analysis** | R DESeq2 GSEA |
A transcriptomics workflow for differential gene-expression analysis and gene set enrichment in Crohn's disease data. |
| ** QSAR Prediction with RDKit & SHAP** | Python RDKit scikit-learn SHAP |
Molecular descriptor and fingerprint analysis, predictive QSAR modeling, model evaluation, and interpretation of predictions. |
| ** hERG Pharmacophore Modeling** | MOE Pharmacophore Virtual Screening |
A computational modeling study examining pharmacophore features and screening approaches for hERG-related compounds. |
| ** Bioinformatics with Python** | Python Biopython Sequence Analysis |
Educational scripts covering FASTA/FASTQ processing, biological sequence analysis, and introductory bioinformatics workflows. |
These projects include independent computational studies and educational work; they are not presented as peer-reviewed publications. Their GitHub repositories are currently private and will be linked individually as they are prepared for public release.
Further work and research figures: Computational Projects
The diagram below summarizes my main research interests and the computational methods I work with.
mindmap
root((Hazrat Maghaz
Research Focus))
Bioinformatics
RNA-seq analysis
Sequence analysis
Metagenomics
Reproducible workflows
Biomedical AI
NeuroBridge
Computational pathology
Machine learning
Multimodal analysis
Genomics
Gene-expression profiling
Differential expression
Pathway enrichment
Biological interpretation
Drug Discovery
QSAR modeling
Molecular docking
Molecular dynamics
Pharmacophore modeling
Research Software
Python and R pipelines
FastAPI and Next.js tools
Data visualization
Scientific reporting
- MS Bioinformatics, NUST SINES (2024–2026). NeuroBridge thesis approved in September 2026.
- Research Assistant, NUST SINES (2025–2026). Computational work involving histopathology images, gene-expression data, and machine-learning analysis.
- BS Biochemistry, Abdul Wali Khan University Mardan (2020–2024). Undergraduate thesis investigating potential HIV-1 integrase inhibitors using QSAR analysis and molecular docking.
- CD28 structural bioinformatics study, manuscript submitted in 2026. Acceptance or publication has not been confirmed.
I welcome discussions about PhD research opportunities, academic collaborations, and computational projects in bioinformatics, biomedical AI, genomics, transcriptomics, and computational drug discovery.
