Services
End-to-end computational biology support
For teams from preclinical biotechs to large pharma. I work across cardiovascular, fibrosis, immunology, and neuroscience, from discovery to the clinic and back.
Multi-Omics Data Analysis
Comprehensive analysis of high-dimensional biological datasets, including RNA-seq, single-cell RNA-seq, proteomics, and integrated multi-omics approaches. From experimental design through to publication-ready results.
- Bulk and single-cell RNA-seq analysis
- Spatial transcriptomics
- Proteomics: affinity platforms (Olink, SomaScan) and mass-spec data
- Multi-omics data integration
AI for Data Analysis
Practical LLM tooling that turns data and knowledge into decision-ready insights. It speeds up analysis and evidence-gathering, with the guardrails to make it trustworthy.
- LLM-accelerated analysis of RNA-seq, proteomics, clinical-trial, and real-world data
- Knowledge discovery from biomedical literature, internal documents, and patents
- Custom LLM tools and workflows tailored to your team’s data and questions
- Checking LLM outputs against ground truth, with accuracy benchmarks and hallucination detection
Related notes: testing a single-cell foundation model against a PCA baseline · a PubMed search tool for Claude
In development: TensorTarget, an AI co-scientist for drug-target hypotheses →
Targets, Biomarkers & Patient Stratification
Human genetics and multi-omics evidence for the decisions that shape a program (which target, which indication, which patients), grounded in causal evidence, in both directions between discovery and the clinic.
- Target identification and validation from human genetics and multi-omics evidence
- Indication expansion: where a target’s biology applies across diseases
- Biomarker discovery from patient and preclinical data
- Biomarker panel design for affinity platforms (Olink, SomaScan)
- Patient stratification for trials and programs
- Post-hoc analysis of clinical trial data: responder biomarkers, subgroups, and pharmacodynamic response
- GWAS, PheWAS, eQTL, pQTL, and biobank-scale genetics (UK Biobank and similar)
Related notes: rereading two failed TYK2 trials in Crohn’s disease · a Bayesian go/no-go for a drug candidate
Related TheraRadar briefs: how often cancer accelerated approvals are confirmed · reading a cancer trial’s endpoints · IL-23 drugs in Crohn’s disease
TheraRadar heatmaps of drugs by indication: JAK/TYK2 inhibitors · IL-23 axis inhibitors
Bioinformatics Pipeline Development
Custom bioinformatics pipelines for any omics technology (NGS, proteomics, spatial transcriptomics, single-cell, metabolomics), built for reproducibility, scalability, and integration with your existing infrastructure.
- Pipeline development across NGS, proteomics, spatial, single-cell, and other omics modalities
- Custom algorithm development for sequencing and high-dimensional omics data
- Pipeline optimization and modernization
- Reproducible & auditable analysis that makes your computational work reviewer-ready and audit-traceable
Not on this list?
If your question doesn't fit the four items above, that's usually a sign it's the right kind of problem for a conversation. Talk to me →
Confidentiality & data handling
Whatever the service, every engagement runs on the same footing: your data and IP stay yours.
- NDA-first: a mutual NDA before any data or program detail changes hands.
- Nothing client-identifying is ever published. No client names, logos, targets, compounds, or data appear here or anywhere without your written permission.
- Your data stays in your environment where possible; otherwise it is isolated, encrypted, and deleted when the engagement ends.
- AI and your data: I use your data only for your project, with AI tools that have model training turned off. If your team has its own rules for AI, such as approved tools only or no AI at all, I’ll work within them.
Have a project in mind?
Every engagement starts with a conversation. Tell me what you're trying to solve.
Get in Touch