Start with the decision or experiment
Define what will be chosen, tested, or changed. Then decide what evidence would be sufficient to act.
James Young / scientific machine learning
Machine learning for biological discovery and operational decisions.
I build analytical systems that narrow complex search spaces, prioritize experiments and interventions, and translate data into decisions.
01 / Perspective
Scientific depth, analytical judgment, applied leadership.
PhD in Biological Sciences
Specializing in molecular biology, with work spanning functional discovery, comparative genomics, multi-omics, and protein learning.
Peer-reviewed computational biology
Published work on oxic nitrogen fixation and broader biological data integration.
Applied machine-learning leadership
Forecasting, causal inference, targeting, resource allocation, and implementation.
02 / Selected work
Questions where search, evidence, and decisions meet.
The work connects research questions and operating constraints to analytical choices and concrete outcomes.
Biological discovery 01
Multi-omic discovery for oxic nitrogen fixation
Which undercharacterized genes should be prioritized for experiments on oxygen-tolerant nitrogen fixation?
View work Biological discovery 02
Protein-language-model search for Rubisco variants
How can a large protein sequence space be narrowed to a small, informative set of variants for experimental consideration?
View work Decision systems 03
Operating principles for decision systems
What makes an analytical system usable in recurring operational decisions?
View work 03 / Open research
Recent machine-learning publications.
Recent papers span gene-function discovery, protein representation, and metabolite-based classification.
Predicting FOX gene candidates for oxic nitrogen fixation using multi-omic machine learning and comparative bioinformatics
Open paper
Active Learning on Protein Language Model Embeddings Accelerates Rubisco Variant Discovery for Desired Traits
Open paper
Secondary Metabolites Predict Diazotrophic Cyanobacteria: A Model-Based Cheminformatic Approach
Open paper
04 / Current research
ML-guided functional discovery in undercharacterized microbial systems.
Oxygen-tolerant nitrogen fixation in cyanobacteria is the flagship application within a broader research program.
The broader method connects comparative genomics, condition-specific multi-omics, protein representations, candidate prioritization, active learning, and experimental validation planning.
01Define the biological decision
02Integrate sequence, context, and multi-omic evidence
03Rank candidates with uncertainty visible
04Plan validation and update the search
05 / Approach
The model is one part of the reasoning system.
Constrain the search space
Use scientific context, operating constraints, and prior evidence before adding model complexity.
Design for the next cycle
Make validation, interpretation, adoption, and iteration part of the system from the beginning.