Direkt zum InhaltDirekt zur SucheDirekt zur Navigation
▼ Zielgruppen ▼

Humboldt-Universität zu Berlin - IRI Life Sciences

Future events

  • 2019-10-24T16:00:00+02:00
  • 2019-10-24T17:00:00+02:00
Th 24. Oct.

Oct 24, 2019 from 04:00 PM to 05:00 PM
IRI Life Sciences, Philippstr. 13, Building 18, 3rd Floor, Room 410

Abstract: Improving our understanding of gene regulation in the context of development and diseases, such as cancer, is one of promises of the large data collection efforts currently underway. However, the complexity of gene regulation makes it anything but trivial to condense information from large amounts of sequencing data into testable biological hypotheses. I will present two novel computational approaches toward that goal. First, I will describe a novel machine learning approach for learning, possibly distal, transcriptional regulatory elements for gene expression from paired epigenomics and transcriptomics data. I will delineate the mathematical formulation and our experimental results for their validation. Second, I will present a multitasking-based optimization approach that can be used to learn regulatory associations for regulators, such as microRNAs and transcription factors, on the level of individual transcript isoforms. I will present results that illustrate that transcript-isoform specific (mis)regulation may be an important regulatory switch in cancer cells.