Meet the speaker
If you want to meet Dr. Jacob Schreiber at Forskningsparken on Tuesday June 17th, fill in the jotform at https://form.jotform.com/251603757511353.
Abstract
Sequence-based machine learning (ML) models have proven invaluable for studying regulatory genomics, with two key application areas being identifying the sequence features driving biochemical activity and designing novel sequences that exhibit desired characteristics. However, these seemingly different tasks are just two sides of the same coin, with the algorithms for each being remarkably similar. To demonstrate this, I will first introduce DragoNNFruit, a dynamic neural network for identifying and extracting the continuous regulatory drivers of chromatin accessibility at single-cell resolution. Then, I will introduce Ledidi, which turns any already-trained ML method into a designer of genomic sequences with desired characteristics.