BioNTech Investor Day Presentation Deck slide image

BioNTech Investor Day Presentation Deck

2 COVID-19 variants monitoring and prediction Reduction in time to detect new variants of concern by ~2 months SARS-CoV-2 virus www Structural modeling Spike protein ACE2 SHEE Epitope alteration score Detection and neutralization by antibodies ACE2 binding Score Spike protein binding to ACE2 receptor EWS score Immune escape score Pareto score Infectivity score Machine learning modeling Semantic change ¹ Artificial intelligence collaboration of BioNTech and InstaDeep. EWS, emergency warning system. Beguir K, et al. bioRxiv 2021; doi: 10.1101/2021.12.24.474095. Distance to wild-type in embedding space Log-likelihood Sum of predicted probabilities over the residues Metadata (sequences + experimental data) Language model Growth rate Cumulative sum of variant cases over time (in log scale) Beta (B.1.351) 10k 1000 100 10 1 10k 1000 100 10 1 100 10 1 Jul 2020 Jan 2021 Jul 2021 Digital & AI/ML Epsilon (B.1.427, B.1.429) Jul 2020 Jan 2021 Jul 2021 Jul 2020 Early computational detection¹ of high-risk SARS-CoV-2 variants supports rapid COVID-19 vaccine adaptation to combat new threats, saving months in response time Theta (P.3) Jan 2021 Jul 2021 Flagging by EWS WHO designation InstaDeep™ BIONTECH 43
View entire presentation