Abstract
Viral host-switching from host H1 to host H2 is often associated with changes in viral evolutionary rate r. The pre-switching rate r1 in H1 may stay the same or increase/decrease to a new rate r2 in H2 during the host-switching and host-adapting process, depending on the difference between H1 and H2. The changing rate has previously been modeled by a linear function when the time interval is short but is better modeled by a sigmoidal function. The author presents the mathematical model, illustrates its application, and implements the rooting and dating methods in a new version of the user-friendly TRAD program, which is freely available at https://dambe.bio.uottawa.ca/TRAD/TRAD.aspx. Application of the method to a phylogeny of early SARS-CoV-2 genomes revealed (i) an increase in r in late February 2020 contributed mainly by the D614G lineage, (ii) a significantly better fit of the sigmoidal-rate model to the SARS-CoV-2 evolution than the constant-rate model, and (iii) the common ancestor of the included SARS-CoV-2 genomes dated to November 20, 2019.