Deep Learning for Analysis and Integration of Large Biological Data

About the Talk

Presenter

Tunca Doğan: KanSiL, Department of Health Informatics, METU Graduate School of Informatics, and the European Bioinformatics Institute (EMBL-EBI)

Abstract

Machine learning and data mining are used to interpret large biological and biomedical datasets, particularly in predicting protein properties such as functions and interactions.

This talk covers two deep learning methods developed at our research centre: DEEPred, for predicting the functions of uncharacterised proteins, and DEEPscreen, for identifying novel drug candidate molecules targeting the human proteome. Alongside the benefits of deep learning approaches for biological data, it also addresses their limitations: deep learning alone cannot solve every problem related to bioinformatics.

Date: August 2nd, 2018 – 14:00 (London)