T4 – Deep learning for predicting protein–DNA and –RNA binding
Date: September 8, 2018
Time: 9:00 – 17:00
Yaron Orenstein, Ben-Gurion University
In this tutorial, we will cover the basic building blocks of deep neural networks, both convolutional and recurrent, emphasizing their relevance in recognizing DNA and RNA patterns. We will discuss typical datasets in the field of protein–DNA and –RNA binding and focus on how deep neural networks can be trained on them to produce accurate predictive binding models. We will conclude with introductory programming experience with Keras package for the purpose of learning protein binding preferences from experimental data.
Undergrad and grad students, staff, researchers, post-docs and investigators are all welcome. Background in bioinformatics, mathematics, machine learning and programming is assumed.
Participants are encouraged to bring their laptops, with Python and Keras package installed on it or on a computing server they have access to.
9:00-10:30 Key challenges in the field of protein–DNA and –RNA binding
11:00-12:30 Basic building blocks in deep neural networks, both convolutional and recurrent.
13:30-15:00 Applying neural networks to DNA/RNA sequences and train them.
15:30-17:00 Practical programming examples using Keras.