Introduction

This workflow processes paired-end Illumina SARS-CoV-2 RBD amplicon sequencing data collected from air samples in Ramuta et al. 2022, SARS-CoV-2 and other respiratory pathogens are detected in continuous air samples from congregate settings.

Getting started

This workflow uses the Snakemake workflow manager. If you don't have snakemake installed on your computer, you need to follow five steps:

1. Install the miniconda python distribution (https://docs.conda.io/en/latest/miniconda.html)
2. Install the mamba package installation tool: conda install -y -c conda-forge mamba
3. Install snakemake into a new python virtual environment: mamba create -c conda-forge -c bioconda -n snakemake snakemake
4. Activate the new environment: conda activate snakemake

Try typing snakemake -v. If your terminal returns something like 6.12.3 snakemake is installed correctly.

Bundled data

• SARS-CoV-2 RBD amplicon BED files from Dr. Marc Johnson are in the resources/amplicon folder.
• SARS-CoV-2 RBD primer scheme from Dr. Marc Johnson is in the resources/primers folder.
• An example SARS-CoV-2 genome reference in the resources/genomes folder.

Workflow summary

• This workflow supports multiple interleaved FASTQ file inputs
• Paired-end reads are merged into synthetic reads spanning the entire RBD PCR amplicon
• Synthetic reads are mapped to a viral reference sequence
• Mapped read positions are compared to a BED file of amplicon sequences. The BED file shows amplicon coordinates after PCR primer trimming, representing the "full-length" amplicon sequences. Synthetic reads that span the entire amplicon are retained.
• These reads are downsampled on a per-amplicon basis
• The downsampled reads are mapped to a viral reference sequence and a consensus sequence is generated. The mapping BAM file can be loaded into Geneious or other tools for variant calling.
• Downsampled reads are deduplicated and those deduplicated clusters that exceed a minimum frequency are returned as putative haplotype sequences. These can be used in downstream analyses to determine linkage of variants within the same amplicon.

Configuration parameters

• reads - path to interleaved Illumina FASTQ files (expects .gz compressed FASTQ)
• amplicon_bed - path to BED file that contains one row per amplicon, with coordinates of amplicon relative to reference after primers have been removed.
• primer_bed - path to BED file that contains one row per primer pair.
• ref_fasta - path to FASTA file with viral reference genome
• target_depth - number of reads to retain from each amplicon. For example, '1000' would retain a maximum of 1,000 reads from each amplicon specified in amplicon_bed
• output_prefix - string to use as the prefix for all output files
• consensus_min_depth - when generating a consensus sequence, the minimum coverage required for a consensus basecall. For example, if this is set to '20' sites that have coverage of '12' would be masked with an 'N'
• haplotype_min_frequency - frequency of deduplicated reads to report in haplotypes. For example, setting this to '0.01' and 'target_depth' to 1000 would retain sequences that are found in at least 10 identical downsampled reads (1%). Note that this expects that there are 'target_depth' reads for each amplicon; amplicons with unexpectedly low coverage will still report sequences found in at least 10 identical reads.

Configuration parameters used in the manuscript

• reads - 'data/'
• amplicon_bed - 'resources/amplicons/SARS-CoV-2.MJ-RBD-NTD.amplicon.bed'
• primer_bed - 'resources/primers/SARS-CoV-2.MJ-RBD-NTD.primer.bed'
• ref_fasta - 'resources/genomes/MN908947.3.fa'
• target_depth - '1000'
• consensus_min_depth - '20'
• haplotype_min_frequency - '0.001'

Output files

• [output_prefix].[target_depth]X.fastq.gz - FASTQ files after downsampling to target depth
• [output_prefix].[target_depth]X.primer_removed.sorted.bam - BAM file after mapping downsampled FASTQ to ref_fasta and removing primers specified in primer_bed
• [output_prefix].[target_depth]X.primer_removed.sorted.bam.bai - BAM index of the mapped file
• [output_prefix].[target_depth]X..primer_removed.consensus.fa - FASTA file with consensus sequence derived from BAM file
• [output_prefix].[target_depth]X..primer_removed.consensus.qual.txt - Text file with consensus sequence quality scores derived from BAM file
• [output_prefix].[target_depth]X.haplotypes.fastq - FASTQ file where identical reads at a frequency greater than haplotype_min_frequency are retained. Useful for mapping to a reference, extracting all reads corresponding to a single amplicon, and identifying sequences corresponding to haplotypes.

Manuscript data

Interleaved sequencing data for the 9 air samples included in Ramuta et al. 2022 have been deposited in Sequence Read Archive (SRA) under bioproject PRJNA811594 and should be downloaded into the data/ directory to replicate the analysis in the manuscript.

Analysis

Run snakemake with: snakemake --use-conda -- cores 8

For more information

Contact Dr. David O'Connor (dhoconno@wisc.edu), Dr. Shelby O'Connor (slfeinberg@wisc.edu), or Mitchell Ramuta (ramuta2@wisc.edu)