Mirza M.

I turn complex biological data into deployable models, published frameworks, and results you can act on - across drug discovery, multi-omics, genomics, and clinical ML. I've built a deep learning framework for ADC linker design (ROC-AUC 0.878), a transformer-based spatial transcriptomics system applied to human brain tissue, a clinical risk model deployed as a live web calculator (AUROC 0.827, n=7,225 patients), and a ctDNA aptamer prediction pipeline for liquid biopsy. All delivered to publication standard for international clients. What I work on: - Drug discovery - molecular docking, MD simulation, virtual screening, drug repurposing, ADC linker optimization, multi-epitope vaccine design - Genomics & NGS - WGS/WES, Oxford Nanopore long-read, variant calling (GATK, Clair3, DeepVariant), PRS analysis, population genomics, Sanger sequencing & GenBank submission - Transcriptomics - bulk RNA-seq (HISAT2, DESeq2), scRNA-seq (Seurat, Scanpy), spatial transcriptomics, multi-omics integration - Machine learning - XGBoost, deep learning, graph neural networks, transformers, clinical prediction & risk modeling - Diagnostics - TaqMan multiplex qPCR, digital PCR, primer & probe design, aptamer discovery - Synthetic biology - DNA origami nanocarriers, codon optimization, recombinant construct design - Microbiome - QIIME2, DADA2, MetaPhlAn, ANCOM-BC, ALDEx2 - Cloud & infrastructure - AWS SageMaker, reproducible pipelines, Python, R, Bash, Linux If your project involves biological data of any kind, I'll build the pipeline, run the analysis, and deliver something you can publish, present, or deploy. Clean code. Reproducible pipelines. Publication-ready figures. 5–10 day turnaround.