G-quadruplex forming sequences in the genome of all known human viruses: a comprehensive guide

Enrico Lavezzo; Michele Berselli; Ilaria Frasson; Rosalba Perrone; Giorgio Palù; Alessandra R. Brazzale; Sara N. Richter; Stefano Toppo

The copyright holder for this preprint (which was not peer-reviewed) is the . https://doi.org/10.1101/344127 doi: bioRxiv preprint means that the median G4 content in real viral sequences is not significant if compared to a random distribution; conversely, a P-value close to 0 means that G4 content is highly significant. This interpretation holds independently of the length of the genome and/or of the prevalence of either G/C bases or G-islands, as we compare the number of G4s in a viral genome with the one we would expect in a simulated genome of the same length and of either the same base or G-island composition. To account for possible high discreteness of the data, a less conservative version of the P-value, called the mid-P value (Armitage and Berry 1994) , was used. Segment diagrams of the mid-P values of the Baltimore grouped viruses are reported in Supplemental_Material.pdf (Figure S1a and Figure S1b ) (Chambers et al. 1983 ). The number of viruses whose median G4 count is significant at the 10% level is listed in Table 2 (virus names in Supplemental_Table_S3.xlsx) with the indication of whether this median count is either higher or lower than the G4 count in simulated sequences. Our data show that most members of the dsDNA, ssDNA, and ssRNA (RT) present a highly significant content of G4s formed by GG-, GGG-and/or GGGG-islands in one or both strands. ssRNA (-) and ssRNA (+) classes are heterogeneous since some viruses are highly significant in any G4 category (from GG-to GGGG-islands), while others are not (see below). Formation of G4s in members of the dsRNA group is notably less significant.