Cryptic Variant Detection

SARS-CoV-2 wastewater cryptic variant detection pipeline.

Installation

Install via Git

git clone https://github.com/dylanpilz/cryptic-variants.git
cd cryptic-variants

Environment setup

Via conda:

conda env create -f environment.yml
conda activate cryptic-variants

Via mamba:

mamba create -n cryptic-variants
mamba env update -n cryptic-variants --file environment.yml
mamba activate cryptic-variants

Usage

GISAID Authentication

To access data from GISAID, you must have a GISAID account. In order to authenticate your credentials, run the following command:

python -c "from outbreak_data import authenticate_user; authenticate_user.authenticate_new_user()"

This will provide a link prompting you to enter your GISAID username and password. Once completed, your token will be saved and you will not need to authenticate again.

Running the pipeline

nextflow run main.nf -entry [from_fastq|from_bam] --input_dir <path/to/input/dir>  --output_dir <path/to/output/dir>

Set -entry to from_fastq if you are providing paired fastq files, or from_bam if you are providing aligned bam files. In the output directory, you will find a cryptic_variants directory containing potential cryptic variants, as well as a covariants directory containing the output from freyja covariants for the provided samples (see freyja).

Optional parameters

--ref <path/to/reference.fasta>
            Reference genome to use for alignment and covariant detection
            (default: data/NC_045512.2_Hu-1.fasta)
--gff_file <path/to/gff_file.gff>
            GFF file containing gene annotations
            (default: data/NC_045512.2_Hu-1.gff)
--primer_bed <path/to/primer.bed>
            BED file containing primer locations for primer trimming
            (default: data/nCoV-2019_v3.primer.bed)
--skip_trimming <true|false>
            Whether or not to trim primer sequences from reads. If true, primer
            trimming will be skipped.
            (default: false)
--min_site <int>
            Minimum genomic site to consider for cryptic variant detection
            (default: 22556) (RBD start)
--max_site <int>
            Maximum genomic site to consider for cryptic variant detection
            (default: 23156) (RBD end)
--min_WW_count <int>
            Minimum number of wastewater hits to consider a cluster of
            variants in a given sample
            (default: 30)
--max_clinical_count <int>
            Maximum number of clinical hits for a variant to be considered
            cryptic
            (default: 5)
--location_id <str>
            Location ID to query from GISAID
            (default: 'global')

Output

The pipeline produces two output directories: cryptic_variants and covariants. The cryptic_variants directory will contain a {sample}.covariants.cryptic.tsv for each sample in the input directory. This file contains the following columns:

Covariants Mutation cluster detected WW_Count Number of wastewater hits for the mutation cluster in this sample Clinical_Count Number of clinical hits for the mutation cluster in this sample Lineages Lineages associated with the mutation cluster (if any)

Covariants      WW_Count        Clinical_Count  Lineages
['S:G416E', 'S:K417N', 'S:N440K', 'S:L452Q']    12      0       NA
['S:K417N', 'S:S438P', 'S:N440K', 'S:L452Q']    14      0       NA
['S:K417N', 'S:Y421H', 'S:N440K', 'S:L452Q']    11      0       NA
['S:G416G', 'S:K417N', 'S:N440K', 'S:L452Q']    11      2       ['ba.2.12.1']
['S:K417N', 'S:N440K', 'S:G446G', 'S:L452Q']    14      2       ['bg.5']
['S:K417N', 'S:N440K', 'S:L452Q', 'S:F456F']    15      2       ['ba.2.12.1']

The covariants directory contains the raw output from freyja covariants.

andersen-lab/Cryptic-Variants