Communications in Information and Systems

Volume 21 (2021)

Number 1

SARS-CoV-2 becoming more infectious as revealed by algebraic topology and deep learning

Pages: 31 – 36

DOI: https://dx.doi.org/10.4310/CIS.2021.v21.n1.a2

Authors

Jiahui Chen (Department of Mathematics, Michigan State University, East Lansing, Mich., U.S.A.)

Rui Wang (Department of Mathematics, Michigan State University, East Lansing, Mich., U.S.A.)

Guo-Wei Wei (Department of Mathematics, Michigan State University, East Lansing, Mich., U.S.A.)

Abstract

Severe acute respiratory syndrome coronavirus 2 (SARS‑CoV‑2) caused by coronavirus disease 2019 (COVID-19) has led to a tremendous human fatality and economic loss. SARS‑CoV‑2 infectivity is a key reason for the widespread viral transmission, but its rigorous experimental measurement is essentially impossible due to the ongoing genome evolution around the world. We show that artificial intelligence (AI) and algebraic topology (AT) offer an accurate and efficient alternative to the experimental determination of viral infectivity. AI and AT analysis indicates that the on-going mutations make SARS‑CoV‑2 more infectious.

Keywords

viral infectivity, binding affinity change, mutation, deep learning, persistent homology

2010 Mathematics Subject Classification

Primary 68T01, 68U01, 92B05. Secondary 00A69.

The full text of this article is unavailable through your IP address: 3.136.26.20

This work was supported in part by NSF Grants DMS-1721024, DMS-1761320, and IIS1900473, NIH grant GM126189, Bristol-Myers Squibb andPfizer.

Received 7 November 2020

Published 8 February 2021