(A) Main features of single-use collection pad. The filter paper containing the attractor is secured behind the saliva collecting filter paper, with a protective grid in between. (B) Experimental design of this work. (C) Saliva collector being licked by laboratory mouse and shelter cat.

Predictors of variation in viral Cq values.

A) Rabies variation was explained by day, with viral Cq increasing non-lineary with day; B) HSV2 and C) Zika values were explained by an interaction between temperature treatment and day, with Cq value increasing with day only under high humidity hot (HHH) conditions.

Variation in beta-actin gene Cq values from licked paper obtained from collector devices installed in mouse cages during experimental rabies infection with the A) AgV3 and B) CVS RABV strains.

In both cases, beta-actin was detected on each day of the experiment, but levels of detection varied across days. Samples were run in duplicate on each day.

Comparison of FeLV paired Cq values from oral swabs and licked paper for 13 cats.

Viral Cq values were significantly higher for the paper samples.

Taxonomic composition of mice saliva samples over the course of rabies infection.

Proportion of each most frequent species found across mice samples based on the hits detected in each sample. (Upper panels) Beneficial bacteria of the mouse microbiome. (Middle panels) Mice-specific pathogens. (Lower panels) Chow components. All values are represented as 95% CI (Wilson method).

Taxonomic composition of cats saliva samples.

Proportion of each most frequent species found across cats samples based on the hits detected in each sample. All values are represented as 95% CI (Wilson method). A) Feline-specific pathogens. The symbol # indicates a SARS-CoV-positive cat. B) Relative frequency of Conchiformibius kuhniae and C. steedae, beneficial bacteria of the cat’s oral microbiome (upper panel) and of uncultured organisms (lower panel).

The participating shelter had a maximum occupancy of 40 cats in individual/family dens.

A) Individual (right side) and family (left side) dens. B) Intensive care box.

Tapestation results for cat RNA samples (sequencing run 2).

Nanopore MinION sequencing run reports.

The vertical dotted grey line in the lower panel indicates 16hs (run 2) for comparison with the upper panel (run 1, total of 16hs of run).

Results of linear mixed effects models with stepwise elimination examining predictors of variation in the Cq values of multiple viruses (Rabies, HSV1, HSV2, Zika).

Terms included in the best fitting model are listed with significant terms in bold.

Post hoc tests for HSV2 linear mixed effects model.

Significant terms are listed in bold.

Post hoc tests for Zika virus linear mixed effects model.

Significant terms are listed in bold.

Individual Cq values from cat samples.

All cats were tested for presence of FeLV virus and FIV antibodies by analysis of blood (jugular venipuncture) by SNAP test (Idexx™) at admittance in the shelter (“serological status”). RTqPCR of saliva (oral swab or spontaneous licking samples) for FeLV and/or FIV or Nanopore™ metabarcoding (of spontaneous licking samples) were performed and data is listed in the table. Numerical value = mean Cq of RTqPCR duplicates. nd = not done. ambiguous = inconclusive result (one replicate did not amplify OR duplicates amplified only one product of wrong Tm). Yellow highlight = sequencing run 1. Pink highlight = sequencing run 2.

Results of linear mixed effects model with stepwise elimination examining predictors of variation in Zika virus Cq values (outlier included).

Terms included in the best fitting model are listed with significant terms in bold.

Post hoc tests for Zika virus linear mixed effects model with outlier included.

Significant terms are listed in bold.