Internal Seminar: Masai Unit and Kitano Unit

Join us for this February's Internal Seminar Series, from 17:00 to 18:00 in B250, central building.

This month's seminars feature the Developmental Neurobiology Unit (Ichiro Masai) and the Open Biology Unit (Hiroaki Kitano).

Neurobiology Unit (Ichiro Masai)

Speaker : Maria Iribarne

Title : Photoreceptor degeneration in Zebrafish mutant

Abstract : More than two hundred genetic loci that are linked to photoreceptor degeneration were mapped on human chromosomes. However the mechanisms by which photoreceptor degeneration and cell death occur is not fully understood. Since neural circuit structure of the retina is conserved in vertebrate animals from fish to human, we use zebrafish as an animal model. Zebrafish presents several advantages, like rapid development process, and relative short time span from embryo to adult; their embryo and larvae are transparent; external development. We focus on zebrafish mutants showing defects in visual response. In this internal seminar we report a zebrafish mutant showing cone photoreceptor degeneration, namely gold rush (gosh). The gosh mutant is useful for understanding molecular mechanism underlying cone-specific cell death pathway.

Open Biology Unit (Hiroaki Kitano)

Speaker : Kun-Yi Hsin

Title : Computational prediction methods for drug discovery and network pharmacology.

Abstract : Given the rich data and algorithmic resources availability on one side, and urgent needs to capture polypharmacological effects of drugs and candidates on the other side, one of obvious challenges is to develop a computational method that can perform high-precision prediction of drug’s effects over molecular networks. This challenge entails two problems that are: to correctly predict molecular binding interactions, and to apply it over molecular networks to compute aggregated effects of drugs. The concept of systems pharmacology has been hierarchically outlined in recent publications, for which numerous bioinformatics resources across levels from molecule to systems are integrated. In such an integrated framework, molecular interaction is seen a core component that regulates the signaling cascades among molecules. To predict protein-ligand interactions, computational approach like virtual docking simulation technology is beneficial due to its low demand of time and cost to assess the binding potentials of a small molecule to proteins. We aim to develop a novel docking simulation approach using machine learning method to improve the prediction accuracy. Given the significant advancement in capabilities of OIST high-performance computing system, we also attempt to apply molecular dynamics (MD) technology to promote the accuracy of the protein-ligand binding mode after docking. Whilst a promising prediction method is made, we will further deploy the improved docking simulation method to a network-based screening pipeline composed of the facilities of molecular databases and curated molecular interaction network maps. Those maps are disease relevant, such as MAPK or mTOR signaling pathway in cancer research as well as Influenza A Life Cycle pathway map (FluMap). Through the screening pipeline, drug developer can systematically assess the binding potential of a test compound against proteins involved in a particular signaling pathway and to identify the binding targets either primary or off-targets responsible for drug effects and toxicity.