What started out as a set of exploratory, individual steps for analyzing RSV fastq files, has now been amalgamated and integrated into this public repository on GitHub.

The high level architecture of this bioinformatics pipeline is that the inputs are paired-end (PE) fastq sequence files, and the output is a so-called consensus.fasta file for that sample. The consensus.fasta files can be analyzed further (either one at a time, or more commonly concatenated into a multi-fasta file) using Web Applications that will try to add the individual sequences onto an existing phylogenetic tree (e.g. UShER, or Nextstrain/Nextclade).

The major steps use a tool named bwa to align the individual fastq sequencing reads to the RSV reference genome (either RSV-A or RSV-B), which generates a new bam file. We then use ivar to perform some quality control (QC) steps, and also trim (remove) the sequence for the PCR amplication primers which are decorating the ends of each sequenced insert. These steps yield a somewhat smaller version of the bam file. Finally we a different ivar command to create the consensus.fasta file for each sample.

Along the way the pipeline also generates information (reports, or tables) on the type and quality of the sequencing reads, and statistics on how many were correctly aligned to the reference genome. We also calculate the depth of coverage (the read coverage) across the genome, as a measure of success (sequencing success, and alignment success).

A by-product of this pipeline is a list of sequence variants that were detected in the sample, and these variants allow the phylogenetic tree mapping algorithms to assign a patient's sample to a known clade and lineage.