Computational modeling of gene expression based on high-throughput NGS data
Alon Diament
Computational modeling of gene expression mechanisms is crucial to the understanding and the design of every biological system and enables the engineering of synthetic systems containing artificially regulated cellular processes. The presented studies are aimed at developing models of gene expression based on analysis of high-throughput next generation sequencing (NGS) data. Specifically, they focus on: (1) The principles that determine the evolution of the 3D organization of genes in genomes; (2) improving the measurement and modeling of translation and transcript evolution; and (3) applying computational models for the purpose of gene expression engineering.
We have shown, by analyzing high-resolution chromosome conformation capture data (Hi-C) from 5 eukaryotes, that the 3D organization of genes is strongly associated with their function, expression and codon usage (Diament et al., Nat Commun, 2014). In order to study the hypothesis that 3D organization and gene function co-evolve developed a computational approach for studying the organization of multiple organisms in a unified model (Diament and Tuller, Nucleic Acids Res, 2017). Inter-organismal study of the organization of the S. cerevisiae and S. pombe genomes has demonstrated that conserved and diverged modules of spatially organized gene families exist in these genomes, and that such modules are related to conserved and diverged biological processes. Furthermore, we have shown that 3D reconstructions of the S. cerevisiae genome can be improved using predictions that are based on the organization of the S. pombe genome (Diament and Tuller, PLoS Comput Biol, 2015).
Ribosome profiling (or Ribo-seq) is currently the most popular experimental methodology for studying translation; it has been employed in recent years to decipher various fundamental gene expression regulation aspects. In order to understand the resolution and limitations of ribosome profiling (Ribo-seq) we analyzed 15 datasets from 6 organism (Diament and Tuller, Biology Direct, 2016). We further developed an improved protocol for high resolution study of ribosome traffic during translation (Diament et al., PLOS Computational Biology, 2018). Our key finding using this protocol, was that ribosome traffic jams in S. cerevisiae are more frequent than previously thought. Our analysis also suggests that the yeast transcriptome may undergo selection for eliminating traffic jams.
Finally, we have developed algorithms and applied models of gene expression in order to engineer expression and adapt heterologous genes to a host (Diament et al., Submitted, 2018). Thus, the models and algorithms above can be employed to develop future biotechnological and biomedical applications.
העבודה נעשתה בהנחיית פרופ' תמיר טולר , המחלקה להנדסה ביו-רפואית, אוניברסיטת תל-אביב
ההרצאה תתקיים ביום ראשון 18.03.18, בשעה 14:00
בחדר 315, הבניין הרב תחומי, אוניברסיטת תל אביב