Computational Biology and Bioinformatics

Train a sequence model on cytochrome c data, score variants, and test whether it can predict functional protein changes with public datasets.

Build a machine learning model that links satellite heat, tree canopy, and ED visits, then maps environmental justice hotspots.

Build an agent-based model and use smartphone time-lapse photos to quantify SCOBY biofilm growth, gradients, and pattern formation.

Build a Cellular Potts model to test how cytokine gradients and spheroid geometry change immune-cell infiltration and escape routes.

Build an active-learning QSAR model that screens herb-drug interaction risk and measures how well it cuts false negatives in CYP450 prediction.

Train and test histopathology AI models on public cancer tiles, then measure fairness across cohorts while building skills in transfer learning and evaluation.

Build a sequence-based ML model to predict base editing outcomes, then test which nearby bases drive bystander edits with attention maps.

Reconstruct ancestral antifreeze proteins and compare predicted folds to test convergent ice binding with sequence analysis and structure prediction.

Build a Bayesian model for optical illusions, collect web data, and test which priors best explain individual differences in perception.

Build an fMRI classifier that separates imagined from heard speech and maps where the brain signals split using attention analysis.