Article Figures & Data
Figures
- FIG 1
Validation of SourceTracker detection of microbial sequence origins. (A and B) Spike-in experiment in which reads from a single stool sample were added to the reads of palms that otherwise had no fecal signal. The degree to which the added (“spiked-in”) reads were assigned to any fecal sample (A) or the specific stool sample that was spiked in (B) using our SourceTracker protocol is shown. The plots show the results for 20 palms, each with 3 repetitions. Histograms show the count of sample source pairs (e.g., a single stool, oral, or forehead sample) to individual sinks (e.g., a single palm) with percentage mapping separated into self, inside-family, and outside-family sources as measured (C) and with randomly permuted data (D). In each permutation, reads from each palm were randomly assigned across all source samples, and this was repeated 25 times. Error bars represent the standard error measured across these permutations.
- FIG 2
The result of running the SourceTracker analysis on the palms of a single family. The predicted proportions of bacteria on the right (r) and left (l) palms of a family consisting of an infant (1 year old), child (3 years old), mother, and father that are from the forehead (estimated by forehead samples from the same individuals), stool, tongue, and unknown source are shown. Additionally, the hatching pattern indicates the predicted proportion within each of these sources that is specifically from the infant, child, mother, or father.
- FIG 3
Distributions of samples across sources. (A) Percentage of bacteria present on each palm that sources to forehead, stool, tongue, or unknown sources. The forehead, stool, and tongue values are the sum of those mapping to all individual forehead, stool, or tongue samples in the whole cohort. Each point in each column is an individual palm, and each palm is represented in each column. (B) Percentage of bacteria present on each palm that source to forehead, oral, and fecal samples from oneself (the same individual as the sink), members of their family, or unrelated individuals in the study. Each point in each column represents a single palm, and each palm is represented in each column. (C) Average proportions of sequences mapping to oneself, family members, and unrelated individuals in reads that map to forehead, stool, and tongue sources. (D) Average proportions of sequences that map to a given body site based on dog ownership. (E) Average proportions of sequences that map to a given body site from a left or right hand of individuals from a given age group. Error bars show standard error. *, P < 0.05; **, P < 0.01; ****, P < 0.0001.
- FIG 4
Shifts in bacterial origin with age. (A) Average percentage of bacteria present on palms from different body sites broken up by age. (B) Average percentage of bacteria present on palms from oneself, family, and unrelated sites broken up by age. (C) Average percentage of bacteria present on palms from all combinations of body site origins and self, family, or unrelated origins broken up by age. Error bars show standard error. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.
- FIG 5
Gender and parent status affects palm bacterial composition. (A) Average profile of women’s and men’s palms in terms of bacterial site origin. (B) Average proportion of bacteria from origin sites in males and females further broken down by parental status. Only adults are included in this analysis. Error bars represent the standard error of the mean. (C) Average profile of parent and nonparent palms in terms of age group and origin of sample for tracked bacteria. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.
