Learning from pre-pandemic data to forecast viral escape

Thadani, Nicole N.; Gurev, Sarah F.; Notin, Pascal; Youssef, Noor; Rollins, Nathan J; Sander, Chris; Gal, Yarin; Marks, Daniel

doi:10.1101/2022.07.21.501023

Cited by 19 publications

(47 citation statements)

References 109 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, a variety of computational approaches have been developed that attempt to predict escape by new viral variants. These approaches include basic transformations of deep mutational scanning data (Greaney, Starr and Bloom 2022), models that integrate antigenic data with phylogenetic (Neher et al 2016) or sequence data (Sun et al 2013;Harvey et al 2016), and neural networks that can be trained using deep mutational scanning data (Taft et al 2022) or sequence data alone (Hie et al 2021;Thadani et al 2022).…”

Section: Introductionmentioning

confidence: 99%

A biophysical model of viral escape from polyclonal antibodies

Thornton

Hannon

et al. 2022

Preprint

View full text Add to dashboard Cite

A challenge in studying viral immune escape is determining how mutations combine to escape polyclonal antibodies, which can . potentially target multiple distinct viral epitopes. Here we introduce a biophysical model of this process that partitions the total polyclonal antibody activity by epitope, and then quantifies how each viral mutation affects the antibody activity against each epitope. We develop software that can use deep mutational scanning data to infer these properties for polyclonal antibody mixtures. We validate this software using a computationally simulated deep mutational scanning experiment, and demonstrate that it enables the prediction of escape by arbitrary combinations of mutations. The software described in this paper is available at https://jbloomlab.github.io/polyclonal .

show abstract

Section: Introductionmentioning

confidence: 99%

A biophysical model of viral escape from polyclonal antibodies

Thornton

Hannon

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Starr et al used deep mutational scanning to systematically address RBD mutation effects on protein expression, ACE2 interaction, and antibody recognition [12,13]. Recently, a structural investigation found that RBD-ACE2 binding affinity of the Omicron variant is similar to the Delta variant, and that Omicron spike displays significant antibody evasion [14], in line with earlier studies [1,2,15,16]. Ovchinnikov & Karplus used a kinetic model of B-cell affinity maturation to determine the importance of bivalent versus monovalent antibody-antigen interactions in vaccination and infection [17].…”

Section: Introductionmentioning

confidence: 63%

Antibody accessibility determines location of spike surface mutations in SARS-CoV-2 variants

et al. 2023

View full text Add to dashboard Cite

The steady emergence of SARS-CoV-2 variants gives us a real-time view of the interplay between viral evolution and the host immune defense. The spike protein of SARS-CoV-2 is the primary target of antibodies. Here, we show that steric accessibility to antibodies provides a strong predictor of mutation activity in the spike protein of SARS-CoV-2 variants, including Omicron. We introduce an antibody accessibility score (AAS) that accounts for the steric shielding effect of glycans at the surface of spike. We find that high values of the AAS correlate strongly with the sites of mutations in the spike proteins of newly emerging SARS-CoV-2 variants. We use the AAS to assess the escapability of variant spike proteins, i.e., their ability to escape antibody-based immune responses. The high calculated escapability of the Omicron variant BA.5 with respect to both wild-type (WT) vaccination and BA.1 infection is consistent with its rapid spread despite high rates of vaccination and prior infection with earlier variants. We calculated the AAS from structural and molecular dynamics simulation data that were available early in the pandemic, in the spring of 2020. The AAS thus allows us to prospectively assess the ability of variant spike proteins to escape antibody-based immune responses and to pinpoint regions of expected mutation activity in future variants.

show abstract

“…Currently, unsupervised methods trained on natural sequences have been surprisingly successful at learning from evolutionary sequences to predict variant effects. 5,6,[8][9][10][11][12][13][14][15] . Yet, these are often tasks involving single mutations or at single mutational depths.…”

Section: Introductionmentioning

confidence: 99%

Removing bias in sequence models of protein fitness

Shaw,

Spinner,

Shin

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

Unsupervised sequence models for protein fitness have emerged as powerful tools for protein design in order to engineer therapeutics and industrial enzymes, yet they are strongly biased towards potential designs that are close to their training data. This hinders their ability to generate functional sequences that are far away from natural sequences, as is often desired to design new functions. To address this problem, we introduce a de-biasing approach that enables the comparison of protein sequences across mutational depths to overcome the extant sequence similarity bias in natural sequence models. We demonstrate our method’s effectiveness at improving the relative natural sequence model predictions of experimentally measured variant functions across mutational depths. Using case studies proteins with very low functional percentages further away from the wild type, we demonstrate that our method improves the recovery of top-performing variants in these sparsely functional regimes. Our method is generally applicable to any unsupervised fitness prediction model, and for any function for any protein, and can thus easily be incorporated into any computational protein design pipeline. These studies have the potential to develop more efficient and cost-effective computational methods for designing diverse functional proteins and to inform underlying experimental library design to best take advantage of machine learning capabilities.

show abstract

Learning from pre-pandemic data to forecast viral escape

Cited by 19 publications

References 109 publications

A biophysical model of viral escape from polyclonal antibodies

A biophysical model of viral escape from polyclonal antibodies

Antibody accessibility determines location of spike surface mutations in SARS-CoV-2 variants

Removing bias in sequence models of protein fitness

Contact Info

Product

Resources

About