Computational biology + scientific software: Rust/Python tooling, RNA-seq workflows, single-cell analysis, and deployable clinical/product prototypes.

rustscenic (v0.4.4, PyPI): rewriting fragile SCENIC/SCENIC+ regulatory-network workflows as installable, CPU-efficient Rust + PyO3 software, built in collaboration with the Kuan-lin Huang Lab at Icahn Mount Sinai.

Legacy SCENIC stacks need Java, dask, and a long chain of transitive packages that frequently break on current Python; rustscenic ships the practical compute path without them.

• pip install rustscenic; Python 3.10–3.13; Linux/macOS wheels for x86_64 and aarch64
• Five runtime dependencies: numpy, pandas, pyarrow, scipy, anndata
• Core stages in one package: GRN, AUCell, topics, cisTarget, enhancer links, eRegulons
• End-to-end multiome validated on human PBMC and mouse brain E18; stage-level checks across airway, melanoma, and dopaminergic-neuron datasets

• RNA-seq Nextflow pipeline: containerised DSL2 workflow for FASTQ -> QC -> trimming -> HISAT2 -> featureCounts -> DESeq2 -> MultiQC, with Docker/Singularity/AWS Batch profiles, Seqera-ready schema, run reports/traces, and synthetic end-to-end CI.
• Bulk RNA-seq differential expression: DESeq2 on SARS-CoV-2 nasopharyngeal RNA-seq, 1,773 DE genes (n = 60 primary, 99.8% concordant with n = 484 sensitivity), tracked output manifest, full rebuild CI, and Zenodo DOI.
• RustScenic airway validation case study: real-atlas head-to-head against pySCENIC on 31,602 airway cells and 59 regulons, mean per-cell Pearson r = 0.984, 27x AUCell timing difference, CI-tested reference smoke, and Zenodo DOI.
• Airway cell-type deconvolution: PyTorch deconvolution of 484 bulk RNA-seq samples into 14 airway cell types, r = 0.954 on pseudo-bulk 5-fold CV (upper bound), synthetic smoke tests, and model metadata for HVG/cell-type reuse.
• Single-cell immune profiling: Scanpy PBMC pipeline with Scrublet QC, Leiden resolution selection, marker-based annotation, PAGA trajectory inference, T-cell subclustering, full-pipeline CI smoke validation, and generated output checksums.
• SafetyNett: React/TypeScript + Supabase clinical safety-netting prototype from the OpenClaw Clinical Hackathon; CI covers lint, explicit TypeScript checking, production build, and tests.
• scverse: 4 doc improvements to scanpy plotting, fix to PyDESeq2 dataframe handling, open algorithmic PR on AnnData concat API.

evk23umu@uea.ac.uk