Temporal Dynamics, Discovery, and Emergence of Human-Transmissible RNA Viruses

Abstract

Transmissibility, the ability to spread within host populations, is a prerequisite for a pathogen to have epidemic or pandemic potential. Here, we estimate the phylogenies of human infectivity and transmissibility using 1,408 genome sequences from 743 distinct RNA virus species/types in 59 genera. By repeating this analysis using data sets censored by virus discovery date, we explore how temporal changes in the known diversity of RNA viruses-especially recent increases in recognized nonhuman viruses-have altered these phylogenies. Over time, we find significant increases in the proportion of RNA virus genera estimated to have a nonhuman-infective ancestral state, in the fraction of distinct human virus lineages that are purely human-transmissible or strictly zoonotic (compared to mixed lineages), and in the number of human viruses with nearest relatives known not to infect humans. Our results are consistent with viruses that are capable of spreading in human populations commonly emerging from a nonhuman reservoir. This is more likely in lineages that already contain human-transmissible viruses but is rare in lineages that contain only strictly zoonotic viruses.

Keywords: RNA viruses; emergences; epidemic potential; phylogenetics; temporal dynamics.

Fig. 1.

Discovery curves. a) Accumulated numbers of L1, L2, and L34 virus species/types over time. b) As a) showing cumulative fractions. c) Accumulated fractions of virus genera containing L2 or L34 virus species/types or both (note that some genera contribute to both curves). d) Accumulated fractions of virus families containing L2 or L34 virus species/types or both (note that some families contribute to both curves).

Fig. 2.

Bayesian maximum clade credibility (MCC) trees for members of 52 virus genera (excluding 7 single species genera) using polymerase protein sequences (listed in supplementary data file S4, Supplementary Material online). Phylogenies show the most probable transitions between nonhuman viruses (L1), viruses infective to humans (L2), and viruses transmissible in human populations (L34).

Fig. 3.

Temporal patterns. a) Changes in most probable ancestral state (L1, L2, or L34) of human RNA virus genera estimated using sequences from species/types discovered before cutoff date shown. b) As a) showing fractions. c) Changes in the numbers of distinct human-infective RNA virus lineages (human transmissible, strictly zoonotic, or mixed) estimated using sequences from species/types discovered before cutoff date shown. d) As c) showing fractions.

Fig. 4.

Numbers of distinct human-infective RNA virus lineages (n = 149) in different categories. Three lineage types are distinguished: only strictly zoonotic (SZ), only human transmissible (HT), and both (MX). a) Stacked bar chart of strictly zoonotic, human-transmissible, and mixed lineage counts by genus. b) By virus family. c) By enveloped/nonenveloped. d) By virus genome type. e) By vector borne/nonvector borne.

Fig. 5.

Nearest relatives. Matrix showing changes in nearest known congeneric relative from time of discovery to present day for L2 species a) and L34 species b) including 101 species/types and 8 possible categories: L1, L2, L34, L1/L2, L1/L34, L2/L34, L1/L2/L34, and none.