Hazrat Maghaz — Bioinformatics & Computational Biology Consulting

About the work

Biology meets computation.

I am a bioinformatics researcher with an MS in Bioinformatics from NUST SINES. My work spans biological data analysis, machine learning, and computational drug discovery.

My MS research, NeuroBridge, explored the integration of histopathology and gene-expression data for CNS tumor analysis. This academic background informs my computational consulting work.

Explore academic research

01
Start with the research question.
Define the biological objective, expected outputs, and available data before selecting methods.
02
Make methods reproducible.
Prioritize traceable preprocessing, documented analysis, and organized computational workflows.
03
Evaluate results carefully.
Use appropriate comparisons and limitations rather than presenting every computational result as a validated finding.
04
Deliver understandable outputs.
Provide clear figures, method summaries, scripts, and explanations suited to the research team.

Areas of support

Computational expertise.

Focused technical support for studies that need specialized analysis, modeling, or reproducible research code.

Heatmap of transcriptomic measurements grouped by annotated immune-cell populations

01 / TRANSCRIPTOMICS

Genomics & Transcriptomics

RNA-seq workflows, differential expression, sequence processing, enrichment analysis, and biological data interpretation.

Omics

View actual research figure ↗

HES4 UMAP projection with multiple cell clusters, colored by HES4 value

02 / SINGLE-CELL ANALYSIS

Single-Cell Transcriptomics

Single-cell quality control, normalization, embedding, clustering, marker assessment and interpretation of cell populations.

scRNA-seq

View UMAP figure ↗

NeuroBridge multimodal CNS tumor analysis workflow diagram with image and gene-expression branches

03 / MACHINE LEARNING

Biomedical AI & Data Analysis

Machine-learning models, image-feature analysis, multimodal data integration, model evaluation, and interpretation.

AI / ML

View actual research figure ↗

Protein ligand docking pose and two-dimensional interaction visualization from a computational drug design study

04 / MOLECULAR MODELING

Computational Drug Discovery

Molecular docking, QSAR, pharmacophore analysis, structure-based interpretation, and molecular dynamics.

In silico

View actual research figure ↗

MA plot: log fold change versus mean normalized counts, with blue differential features; the source data are not identified as metagenomic

05 / MICROBIOME

Microbiome & Metagenomics

Taxonomic profiling, differential-abundance review, gene-family and pathway interpretation, and biological context for microbial community datasets.

Microbiome

View differential-feature MA plot ↗

06 / SCIENTIFIC COMPUTING

Research Code & Workflows

Python and R analysis pipelines, statistical visualization, workflow documentation, and research software support.

Software

Selected computational work

The methods behind the work.

Selected computational workflows, described from question to approach to deliverable. These examples are not presented as clinical products or peer-reviewed publications.

NeuroBridge — Multimodal CNS Tumor Analysis

Question: Integrating tissue morphology and gene-expression representations for GBM-like versus LGG-like analysis.

Approach: Real WSI patches; ResNet-50 and CTransPath features; RNA embeddings; patient-level evaluation; missing-modality experiments.

Outputs: Classifiers, retrieval examples and methodological comparisons. Research prototype, not a diagnostic system.

Multimodal AI · Computational pathologyRead research overview ↗

Crohn’s Disease RNA-seq Analysis

Question: Connecting differential gene expression with interpretable disease-associated biological pathways.

Approach: RNA-seq processing and QC; DESeq2-based comparisons; enrichment and Hallmark gene-set interpretation.

Outputs: Differential-expression tables, pathway summaries and reproducible figures.

RNA-seq · DESeq2 · GSEAExplore computational projects ↗

QSAR & Explainable Machine Learning

Question: Studying structure–activity relationships from molecular descriptors and predictive modeling.

Approach: Molecular feature preparation with RDKit; train/test evaluation; model comparison and SHAP interpretation.

Outputs: Prediction assessments, descriptor importance and model documentation; study-level claims only.

RDKit · scikit-learn · SHAPExplore computational projects ↗

Molecular Docking & Protein Dynamics

Question: Investigating ligand–protein interactions and structural stability using complementary calculations.

Approach: Structure preparation; docking and interaction review; simulation quality checks; RMSD, RMSF and related trajectory measures.

Outputs: Interaction visualizations, comparative plots and documented simulation methods; not experimental binding evidence.

Docking · GROMACS · Structural bioinformaticsExplore structural projects ↗

Microbiome Functional Interpretation

Question: Linking changes in microbial taxa with gene families and enriched or depleted pathways.

Approach: Taxonomic abundance; functional profiles; differential statistical comparisons; pathway coverage and contextual interpretation.

Outputs: Organized pathway and gene-family tables, explanatory plots and reproducible analysis notes.

Metagenomics · Statistics · Functional profilingExplore computational projects ↗

Structural bioinformatics

Molecules in motion.

Two short examples of molecular visualization based on your supplied clips, optimized for a dark-background portfolio without white video frames.

Trajectory viewer

Protein conformation

Animated three-dimensional representation of a protein structure, showing the molecular geometry over the uploaded visualization frames.

The white canvas was removed for web presentation. These visualizations depict computational structures, not experimental measurements. Use the playback controls to pause or replay.

Original scientific figures