Stereoelectronic-State Informatics
Building representations that treat molecular three-dimensional structure and electronic states as features for explaining reactivity and selectivity.
Research
Research
I combine three-dimensional molecular and electronic-state information with data-driven methods to predict reaction selectivity and identify its physical-organic origins.
Topics
Research Areas
Prediction and Interpretation of Reaction Selectivity
Analyzing how steric and electronic factors control nucleophilic additions to ketones and asymmetric reductions.
Data-Driven Reaction Analysis
Combining machine learning and computational chemistry to derive predictive models and chemically meaningful interpretations from reaction data.
The detailed research page connects the central questions, analytical workflow, and relationships among these themes.
Read research in detailMethods
Computational Chemistry Notes
Technical notes covering molecular preparation, electronic-structure calculations, reaction-path exploration, real-space analysis, cheminformatics, and statistical modeling.
Read the notesCovered Topics
- Electron density, DFT, and basis functions
- Gaussian, GRRM, Multiwfn, and cube workflows
- RDKit molecular preparation and cheminformatics
- Regularized regression, PCA, and steric descriptors
Projects
- Facial selectivity in nucleophilic additions to cyclic ketonesPrediction and interpretation using three-dimensional information. Paper / Press release
- Three-dimensional electronic-state analysis of asymmetric ketone reductionData-driven analysis of selectivity using three-dimensional electronic states. Paper
- Reactivity prediction for oxidative homocoupling of phenolsSubstrate-reactivity prediction using positive and unlabeled machine learning. Paper
Talks and Presentations
- Prediction and interpretation of facial selectivity in ketone nucleophilic reactions using a machine-learning model that evaluates stereoelectronic states, 47th Chemoinformatics Discussion Meeting. Outstanding Presentation Award