Extended Data Fig. 1: Correlation between bacterial load and non-host reads.
Shotgun sequencing was performed on longitudinally sampled intestinal biopsies after processing with host depletion (N = 60 biological replicates). Roughly 25 million reads on average were obtained for each biopsy and all samples fit on a single NovaSeq S1 flowcell. After host-filtering an average of 2 million reads were remaining with a range from 2E4 reads to 2E7 reads. For each box, the middle horizontal line denotes mean values, boxes extend to the 25th and 75th percentile, and whiskers extend to the 1.5 interquartile range. The variability in non-host reads remaining had a strong correlation (Spearman, r = 0.79) with the total microbial load as measured by digital PCR. This strong correlation indicated that our process was achieving a relatively uniform depletion across all samples. Additionally, the strong correlation indicates that the majority of non-human reads in our samples come from bacteria picked up by the 16S primers used for total microbial load quantification.
