Recombinant proteins, clean downstream processes, and the analytics in between.
I work at the bench and in silico on therapeutic protein and peptide candidates — from construct design and expression to chromatographic recovery and the analytical package that decides whether a process is ready to scale.
About
My background sits between molecular biology and process engineering. Most of my work starts with a sequence and ends with a reproducible purification train: expression host, fermentation or transfection format, capture step, polishing, and the analytical readouts that keep each step honest.
I am particularly interested in the unglamorous parts — mass balance, resin lifetime, buffer compatibility, and the small analytical details that separate a process that works once from one that works every week.
Over the past years I have supported projects in recombinant enzymes, peptide hormones, and Fc-fusion formats, mainly at lab and pilot scale. I also build small computational models to prioritise constructs and chromatographic conditions before wet-lab time is spent.
This site is a place for short working notes, method sketches, and occasional reflections on bioprocess practice. It is deliberately low-traffic.
Focus areas
Recombinant therapeutics
Construct design for proteins and peptides, expression host selection, solubility and secretion trade-offs, and early developability assessment.
Downstream processing
Capture and polishing chromatography, buffer and gradient design, resin screening, and scale-up considerations for lab and pilot runs.
Analytical biochemistry
HPLC and LC-MS method development, purity and aggregation readouts, peptide mapping, and practical assay qualification.
Molecular modelling
Structure-based triage of variants, surface and interface analysis, and lightweight in silico screening to guide experimental priorities.
Working notes
When a gradient method travels badly
Method transfer problems are rarely about the column. More often it is dead volume, buffer preparation tolerance, or a detector cell that behaves differently at the new flow rate. I now measure system dispersion before blaming the resin.
Capture step first, elegance later
A modest but robust capture step buys more development time than an ambitious one that only works on a good day. Clarify early, reduce volume, and keep the polishing steps for the questions they are actually good at answering.
Aggregation is a process signal
SEC monomer loss is often treated as an analytical nuisance. In practice it is one of the most useful early indicators that a buffer, a hold time, or an elution pool decision is pushing the molecule in the wrong direction.
Modelling before cloning
A short homology model and a surface map will not predict expression, but they reliably flag constructs that are likely to misfold or stick. Twenty minutes of modelling has saved several weeks of failed batches.
Contact
For method reviews, process questions, or a short conversation about a difficult purification.
[email protected]