We received data in raw peptide count form from 5 Singaporean Eonycteris spelaea bats (raised in captivity) and from 77 Australian Pteropus alecto bats (wild-caught but rehabilitated) after Phage Immunoprecipitation Sequencing (PhIP-Seq) was run in the Wang lab - see subfolders within "/raw-data/PhIP-Seq/" to access these counts. We first sought to convert these peptide counts into Z-scores for input into the AVARDA pipeline (link to repo) for rank categorization of overlapping peptides into putative viral exposures. See here for documentation of the process and pipeline for binning peptide counts off the PhIP-Seq pipeline into Z-scores.
Following establishment of these Z-scores, we input these files into the AVARDA pipeline (link to paper or repo here) to establish putative viral exposures. See "AVARDA" subfolder within the "raw-data" folder for outputs from this pipeline.
After running AVARDA, we merged these viral exposure data with metadata provided by collaborators for all subsequent analyses. The merged AVARDA output data + metadata for both species of bat is available in the "working-data" subfolder as file "merge-data.csv".
We reference these data in all subsequent figure development files.
This repo is organized by figure. Background scripts for generating datasets are found in the "prep-scripts" folder, while all code needed to produce analyses and generate plots for each figure are in the corresponding 'fig1', 'fig2', 'fig3', and 'fig4' subfolders. These folders output files into 'final-figures' and 'supp-figures' accordingly.