Data scientist at Bio-Prodict

I am currently employed at Bio-Prodict as a medior Data Scientist, where I use state of the art machine learning techniques to develop novel solutions for bioinformatics problems.

I am primarily involved in the production of the Helix product. I work in a team that builds on the results of my internships to predict pathogenicity for different protein variants.

Highlights

• Using a variety of machine learning models to create a complex, hypercompetent ensemble model for pathogenicity prediction.
• Implementing research-level neural networks directly from the literature in PyTorch.
• Working with front-end developers to integrate our model's predictions into the user-facing application.
• Adjusting measurement standards to effectively communicate real world performance statistics to users.
• Working to create easy-to-use APIs for internal and external users of our data with documentation.

Skills used

Media

Matthew's Correlation Coefficient performance on an independent dataset of genes, not present in any dataset. Our ensemble outperforms all competitors.