A Protein Language Model for Exploring Viral Fitness Landscapes DOI Creative Commons
Jumpei Ito,

Adam Strange,

Wei Liu

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: March 18, 2024

Abstract Successively emerging SARS-CoV-2 variants lead to repeated epidemic surges through escalated spreading potential (i.e., fitness). Modeling genotype–fitness relationship enables us pinpoint the mutations boosting viral fitness and flag high-risk immediately after their detection. Here, we introduce CoVFit, a protein language model able predict of based solely on spike sequences. CoVFit was trained with data derived from genome surveillance functional mutation related immune evasion. When limited only available before emergence XBB, successfully predicted higher XBB lineage. Fully-trained identified 549 elevation events throughout evolution until late 2023. Furthermore, CoVFit-based simulation JN.1 subvariants Our study provides both insight into landscape novel tool potentially transforming surveillance.

Language: Английский

The Era of the FLips: How Spike Mutations L455F and F456L (and A475V) Are Shaping SARS-CoV-2 Evolution DOI Creative Commons
Daniele Focosi, Pietro Giorgio Spezia,

Federico Gueli

et al.

Viruses, Journal Year: 2023, Volume and Issue: 16(1), P. 3 - 3

Published: Dec. 19, 2023

Convergent evolution of the SARS-CoV-2 Spike protein has been mostly driven by immune escape, in particular escape to viral infection-neutralizing antibodies (nAbs) elicited previous infections and/or vaccinations [...]

Language: Английский

Citations

18
Using big sequencing data to identify chronic SARS-Coronavirus-2 infections DOI Creative Commons
Sheri Harari, Danielle Miller,

Shay Fleishon

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: July 17, 2023

Abstract The evolution of SARS-Coronavirus-2 (SARS-CoV-2) has been characterized by the periodic emergence highly divergent variants, many which may have arisen during chronic infections immunocompromised individuals. Here, we harness a global phylogeny ∼11.7 million SARS-CoV-2 genomes and search for clades composed sequences with identical metadata (location, age, sex) spanning more than 21 days. We postulate that such represent repeated sampling from same chronically infected individual. A set 271 chronic-like was inferred, displayed signatures an elevated rate adaptive evolution, in line validated infections. More 70% mutations present currently circulating variants are found BA.1 predate months, demonstrating predictive nature clades. find probability observing is approximately 10-20 higher transmission chains. next employ language models to most use them infer hundreds additional absence phylogenetic information. Our proposed approach presents innovative method mining extensive sequencing data providing valuable insights into future evolutionary patterns.

Language: Английский

Citations

17
Predicting Functional Conformational Ensembles and Binding Mechanisms of Convergent Evolution for SARS-CoV-2 Spike Omicron Variants Using AlphaFold2 Sequence Scanning Adaptations and Molecular Dynamics Simulations DOI Open Access
Nishank Raisinghani, Mohammed Alshahrani,

Grace Gupta

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: April 3, 2024

Abstract In this study, we combined AlphaFold-based approaches for atomistic modeling of multiple protein states and microsecond molecular simulations to accurately characterize conformational ensembles binding mechanisms convergent evolution the SARS-CoV-2 Spike Omicron variants BA.1, BA.2, BA.2.75, BA.3, BA.4/BA.5 BQ.1.1. We employed validated several different adaptations AlphaFold methodology including introduced randomized full sequence scanning manipulation variations systematically explore dynamics complexes with ACE2 receptor. Microsecond dynamic provide a detailed characterization landscapes thermodynamic stability variant complexes. By integrating predictions from applying statistical confidence metrics can expand identify functional conformations that determine equilibrium ACE2. Conformational RBD-ACE2 obtained using are accurate comparative prediction energetics revealing an excellent agreement experimental data. particular, results demonstrated AlphaFold-generated extended produce energies The study suggested complementarities potential synergies between showing information both methods potentially yield more adequate This provides insights in interplay binding, through acquisition mutational sites may leverage adaptability couplings key energy hotspots optimize affinity enable immune evasion.

Language: Английский

Citations

7
Spatiotemporal dynamics and epidemiological impact of SARS-CoV-2 XBB lineage dissemination in Brazil in 2023 DOI Creative Commons
Ighor Arantes, Marcelo Ferreira da Costa Gomes, Kimihito Ito

et al.

Microbiology Spectrum, Journal Year: 2024, Volume and Issue: 12(3)

Published: Feb. 5, 2024

ABSTRACT The SARS-CoV-2 XBB is a group of highly immune-evasive lineages the Omicron variant concern that emerged by recombining BA.2-descendent and spread worldwide during 2023. In this study, we combine genomic data ( n = 11,065 sequences) with epidemiological severe acute respiratory infection (SARI) cases collected in Brazil between October 2022 July 2023 to reconstruct space-time dynamics epidemiologic impact dissemination country. Our analyses revealed introduction local emergence carrying convergent mutations within Spike protein, especially F486P, F456L, L455F, propelled XBB* Brazil. average relative instantaneous reproduction numbers + F486P F456L L455F were estimated be 1.24, 1.33, 1.48 higher than other co-circulating (mainly BQ.1*/BE*), respectively. Despite such growth advantage, these had reduced on Brazil’s scenario concerning previous subvariants. peak number SARI from wave was approximately 90%, 80%, 70% lower observed BA.1*, BA.5*, BQ.1* waves, These findings multiple progressively increasing yet relatively limited throughout stand out for their heightened transmissibility, warranting close monitoring months ahead. IMPORTANCE one most affected countries pandemic, more 700,000 deaths mid-2023. This study reconstructs virus country first half 2023, period characterized descendants XBB.1, recombinant BA.2 evolved late 2022. analysis supports marked continuous indigenous bearing similar key sites process followed increments without repercussions incidence cases. Thus, results suggest influenced an intricate interplay factors extend beyond virus's transmissibility alone. also underlines need surveillance allows its ever-shifting composition.

Language: Английский

Citations

6
A Protein Language Model for Exploring Viral Fitness Landscapes DOI Creative Commons
Jumpei Ito,

Adam Strange,

Wei Liu

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: March 18, 2024

Abstract Successively emerging SARS-CoV-2 variants lead to repeated epidemic surges through escalated spreading potential (i.e., fitness). Modeling genotype–fitness relationship enables us pinpoint the mutations boosting viral fitness and flag high-risk immediately after their detection. Here, we introduce CoVFit, a protein language model able predict of based solely on spike sequences. CoVFit was trained with data derived from genome surveillance functional mutation related immune evasion. When limited only available before emergence XBB, successfully predicted higher XBB lineage. Fully-trained identified 549 elevation events throughout evolution until late 2023. Furthermore, CoVFit-based simulation JN.1 subvariants Our study provides both insight into landscape novel tool potentially transforming surveillance.

Language: Английский

Citations

6