A method has limited value if it cannot be run, inspected, maintained, and reused.
Scientific software needs to preserve the evidence while making the work operational. My work turns statistical methods, omics workflows, clinical data, and reporting requirements into reproducible tools, structured outputs, and user-facing systems.
Scientific code needs structure, documentation, testing discipline, versioning, and outputs that other people can understand and reuse.
Complex omics and clinical workflows need structured reports, machine-readable outputs, and clear links between input, processing, and interpretation.
Statistical methods become more useful when they are wrapped in usable interfaces, reproducible workflows, and practical decision support.
Biomedical software has to operate across data quality, security, governance, user needs, and long-term maintainability.
Qualifying variant database. The open standard for variant interpretation, with trusted QV sets to enhance clarity and reproducibility in genetics.
QuantCalc. Probabilistic genomic interpretation using priors, observed evidence, and Bayesian inference.
R, Python, Bash, C, Rust, package development, command-line tools, Git, reproducible reports, machine-readable outputs, and software releases.
Linux, Unix, high-performance computing, Docker, Singularity or Apptainer, Nextflow, Snakemake, versioned pipelines, and reproducible analytical frameworks.
SQL, PostgreSQL, EHR-linked workflows, structured metadata, controlled file transfer, encrypted storage, traceability, and data governance.
HTML, CSS, JavaScript, Next.js, React, TypeScript, Supabase, Vercel, APIs, authenticated systems, and user-facing biomedical platforms.
Structured standalone HTML reports, variant interpretation outputs, visual analytics, QC summaries, clinical research dashboards, and evidence-linked documentation.
Machine learning in omics, PyTorch, AI-assisted scientific software, local model deployment, API-based model deployment, and retrieval-augmented scientific workflows.
VCFheader: parsing VCF headers and generating structured standalone HTML reports, described in Lawless (2026).
Archipelago: variant set association visualisation for complex genomic studies, published in Lawless et al (2026).
QuantBayes: Bayesian posterior intervals for genomic variant evidence sufficiency, preprinted in Quant Group et al (2025).
PanelAppRex AI: harmonised disease-gene panels from structured clinical and genetic queries, preprinted in Quant Group et al (2025).
ORCID record:
0000-0001-8496-3725
Application of qualifying variants for genomic analysis. Bioinformatics, 42(2), btaf676, 2026.
Archipelago method for variant set association test statistics. Genetic Epidemiology, 50(1), e70025, 2026.
Built R and Python workflows for genome-wide, rare variant, gene-level, and multi-omic analyses in secure Linux and high-performance computing environments.
WGS, RNA-seq, proteomics, metabolomics, EHR-linked data, approximately 1,000 children, more than 100 TB of biomedical data, structured outputs for clinical research review.
Developed product-facing biomedical data workflows that translate statistical genomics methods into reproducible reporting, probabilistic interpretation, and secure data systems.
R packages, CRAN releases, structured HTML reports, YAML criteria, PostgreSQL, Supabase, Next.js, React, TypeScript, APIs, authenticated scientific software.
Built reproducible statistical and computational workflows across infectious, inflammatory, genomic, and multi-omic cohort studies.
Cohorts up to 5,000 participants, R, Python, Linux, high-dimensional modelling, rare variant analysis, data visualisation, reports, figures, and collaborative scientific outputs.
Biomedical software teams need maintained workflows where data, method, report, and user action stay connected while preserving the underlying science.
