Back to Profiles

 

Job description: I had two overarching tasks. The first was to validate the prior implementation of ChemChainer and DeepChem machine learning tools on the OChem website. I performed this task by reading articles to understand the tools’ implementation by other groups, gathering and preparing the data, running many models on the Ochem website, and comparing our results to the results published by prior implementers.

My second task was to develop a highly accurate model for water solubility. First, I spent considerable time searching the literature for new data and prior solubility modeling tactics. Second, I used multiple methods to analyze the data and determine its’ accuracy. Third, I ran multiple models with the optimized data to get an ideal model for solubility with limited properties. Finally, I explored the effects of the addition of other properties to the solubility model.

This work was largely completed with the OChem website and thus involved little programming. However, I did perform some coding to aid in data processing.

The most personal rewarding part of my internship was gaining knowledge about the computational modeling process and the desire to go back to Stanford to acquire more computational skills.