Date: Wednesday, 15 September 2021

Venue: Cheung Kung Hai Lecture Theatre 2, G/F, William M.W. Mong Block, Faculty of Medicine Building, 21 Sassoon Road, Pokfulam, Hong Kong

Time: 5:00 p.m. - 6:00 p.m.

5:00 p.m.

Speaker: Dr. Xiaolei WANG (Post-doctoral Fellow)
Primary Supervisor: Prof. Jiandong HUANG
Presentation Title: Evaluation of moonlighting proteins as novel vaccine candidates against Staphylococcus aureus infection
Abstract: Accumulating evidence has implied that moonlighting proteins play important roles in pathogenesis of Staphylococcus aureus. Our preliminary data clearly demonstrated that E2 subunit of pyruvate dehydrogenase complex (E2-PDH) of S. aureus can induce robust protection against lethal challenge of S. aureus in murine blood infection models. Immunization of E2-PDH also facilitated the bacterial clearance when mice were subjected to sub-lethal dose of S. aureus. We also found that IL-17A and IFN-γ contributed to the protection induced by E2-PDH in murine skin infection model. In addition to E2-PDH, we continued to search for other moonlighting proteins of S. aureus and four more moonlighting proteins were found to be effective in protecting mice from lethal challenge of S. aureus. Taken together, moonlighting proteins of S. aureus have shown great promise as novel vaccine candidate against S. aureus infection.

5:30 p.m. 

Speaker: Dr. Athena Hoi Yee CHU (Post-doctoral Fellow)
Primary Supervisor: Dr. Alan Siu Lun WONG
Presentation Title: Applying machine learning to combinatorial mutagenesis screens
Abstract: Large-scale pooled library assembly coupled with high-throughput screening can measure functional properties of millions of variants in a single experiment. To further accelerate the search for high-performance variants, we explore the utility of machine learning on predicting the fitness of novel variants based on a small sub-sample of empirical data. Specifically, we apply machine learning to optimise the Cas9 (CRISPR associated protein 9), which can be programmed to cleave specific DNA regions and serves as an important tool for genome-editing. We validated the accuracy of the machine learning models on predicting the activity of Cas9 variants using design libraries. We found that machine learning models can guide the identification of the top Cas9 variants correctly using as little as 10% of training data from the entire library. Our results inform experimental parameters that maximizes the usefulness of machine learning in future screens, including by focusing screening efforts on actively selected mutants and replicate tests on multiple reporter systems.

ALL ARE WELCOME

Should you have any enquiries, please feel free to contact Miss River Wong at 3917 9216.