From (epi-) genomes to transcriptomes: massively parallel inference of regulatory determinants during neural induction
Wednesday July 24 2019 at 11:00
Room 201, Medicine Faculty, TAU
"From (epi-) genomes to transcriptomes: massively parallel inference of regulatory determinants during neural induction"
Dr. Anat Kreimer, Department of Electrical Engineering & Computer Science, Center for Computational Biology, UC Berkeley, Department of Bioengineering and Therapeutic Sciences, Institute for Human Genetics, UCSF
The challenge of mapping genomic regions with functional regulatory role under different conditions, holds a great promise to deciphering the mechanisms that underlie human disorders. I will first present our work to comprehensively map and investigate the function of gene regulatory elements associated with early neural differentiation. As part of this work, I developed a prioritization method that incorporates different genomic data to identify key transcription factors involved in temporal function, several of which were functionally validated to be important and novel neural induction regulators. I will also present computational approaches we developed, for prediction of regulatory regions and the effect of small variants on their regulatory potential. We use these models to understand the robustness of regulatory potential across different cell types. Our results highlight which features are reflective of the cellular context and which are intrinsic to DNA sequence, by exploring the extent of transfer knowledge between cell types. Extending these models to predict functional variants in regulatory elements that are associated with disease, achieves top performance for prioritizing disease relevant variants. My ongoing work focuses on developing predictive models to understand which regulatory elements play a role in specific conditions, cells and tissues, how they interact to achieve transcriptional regulation and what are the mechanisms by which genetic variation in these non-coding regions drives disease in humans.
Dr. Kreimer holds a joint postdoctoral appointment in the Department of Electrical Engineering and Computer Science at UC Berkeley and the Department of Bioengineering and Therapeutic Sciences at UCSF. Her main research interests are focused on developing predictive models of transcriptional regulation by integrating large-scale datasets to shed light on regulatory processes that are condition-specific. She aims to leverage such techniques to identify functional variants and mechanisms of action driving disease. Specifically, she is interested in characterizing the function of gene regulatory elements associated with early neural differentiation and understanding their role in neuropsychiatric disorders - work for which she was granted a K99 award. She holds a PhD in Biomedical Informatics from Columbia University, an MSc in Applied Mathematics and a BSc in Mathematics and Computer Science from Tel Aviv University.