Results
In bats, we obtained on average 40278 (± 3719, n= 8) reads per individual and in birds 71367 (± 26981, n= 10) (Supplementary information: Table S1). From early on, we were able to detect that a large percentage of the reads in the unclamped samples were allocated to a few ASVs and, after performing the taxonomic assignment, we were certain that those reads matched known sequences of chloroplasts and mitochondria from publicly available databases. The chloroplast sequences obtain form the bat samples were a 100% match with chloroplast sequences obtained from chloroplast of Saguaro Columnar cacti (GenBank Accession number: KT164771).
After filtering out the chloroplast and mitochondria assigned reads from the data set, we found that, by using the PNA-DNA clamps, we were able to retain a significantly larger portion of the reads after the filtering step (Fig 4). Although the effectivity varied between individual samples, we always detected an improvement of read coverage available for downstream analyses while using the clamps compared with the unclamped results in pairwise comparisons. On average, the percentage of reads kept improved by 13-fold for the bat (with the clamps cpPNA and mPNA) and by 34-fold for the bird (cPNA and mPNA) (Table S1). The two extreme cases were the bat sample Lepto-195 with a 65-fold improvement and the bird sample MM-143195 with a 216-fold improvement. The control samples, i.e. the bacterial mock community without chloroplasts and mitochondria, showed no fold change indicating that the use of the clamps did not affect the Zymo Mock community in any way (Fig 4). These samples also confirmed that the use of the clamps affected neither the read number nor the percentage of reads kept after the subsequent filtering step.