"CryoEM & BioTech" - 1st session

We have the pleasure to invite to our 1st "CryoEM & BioTech" online seminar:

1st July 2021

1.30pm WEST, Lisbon time / 2.30pm CET 

Guest speaker: Dr. Ilaria Ferlenghi

Head of Structural Microscopy, GSK Vaccines

Zoom link

Meeting ID: 878 6502 7775

Passcode: 397909

An integrated structure-based and ML-based approach for new vaccine identification

Vaccines have proven pivotal against an array of infectious viral and bacterial diseases. However, there are still no effective vaccines against many pathogens as antigen identification and development of vaccine is a time-consuming process.  In this context, finding the most optimal antigen is the bottleneck to make a vaccine accessible to patients everywhere in an accelerated manner.

Several vaccines on the market are composed by antigens expressed on the surface of the microorganism. Identification of these proteins, characterization of their structural insights as well as, their conformation and their antigenic epitopes can accelerate the timeline for the identification and thereafter, the design of novel vaccine(s). 

Towards this, GEMINI, an integrated visual proteomic- based (Cryo-EM, HDxMS, X-ray crystallography) approach in combination with a structure-based similarity search tool for protein screening based on a Structure Similarity Algorithm has been recently developed to allow a rapid identification of new antigens. The GEMINI software is a convolutional neural network-based model that has been trained end-to-end using tm-score as the ground truth for screening structurally similar proteins. This allows the tool to generate score similarity between a new protein and the entire dataset in a fast and antigen focused approach.

Once identified, the novel antigen needs to be evaluated and screened for its functionality and antigenicity before being considered as potential vaccine candidate. An additional GEMINI prediction module tool will allow a structure-based antigenicity scoring where protein structural context as well as physio-chemical protein characteristics are considered to derive the antigenicity-dominant region. GEMINI will be also enriched by ProtLM, a protein functional characterization tool capable to predict protein functionality from their sequence.

Finally, GEMINI, based on advanced structural approaches and important contribution of AI/ML approaches enable a new era of rational structure-based identification of new antigens as potential vaccine candidates.

Meet our guest speaker:

Dr. Ilaria Ferlenghi, Head of Structural Microscopy, GSK Vaccines