Now
Last updated
Working on
- MPhil Computational Biology, Michaelmas term — coursework and problem sets at Cambridge, and settling on a research direction for the year.
- An active-learning harness for perturbation screens — a test bed for comparing acquisition policies (which perturbation to run next) against random and greedy baselines: perturb-policy-harness.
- A PhD application to the Wellcome Sanger Institute — writing the research proposal and talking to prospective supervisors.
Reading
- Perturbation-biology papers — perturbation-response prediction, Perturb-seq analysis methods, and what the recent benchmarks say about which models actually generalise.
- Active-learning papers — acquisition functions and experiment-selection strategies, with an eye on what transfers from the ML literature to real screens.
Looking for
- Research collaborators working on perturbation prediction, active learning for experiment design, or target discovery.
- Early-stage startup people in techbio — founders and operators; I like hearing how methods become companies.