2 Compute high frequency genomic variants

Once your data have been formatted in HaploCoV format (see here), you can derive high frequency genomic variants by computeAF.pl.

The tool partitions genomes according to geographic metadata, and computes allele frequencies over a user defined time interval, in overlapping time windows of a user defined size. The main output consists of lists of high frequency genomic variants at different levels of geographic granularity: global (all genomes), macro-areas (macro geographic areas) and countries (country). Macro areas are according to areaFile in the github repository, see here.

All the final outputs as well as all intermediate files produced by the tool are saved in a folder that can be specified with the option –outdir (see below). Users also have the option to select the minimum frequency (as in allele frequency) and “persistence” (number of weeks above the frequency threshold) thresholds for the identification of high frequency genomic variants.

Options

The script accepts the following parameters:

  • –file file in HaploCoV format;

  • –maxT upper bound in days for the time interval to consider in the analysis (days are counted starting from 12-30-2019). A value of 1 corresponds to 12-31-2019. A value of 365 to 12-30-2020 and so on;

  • –minT lower bound for the time interval. Days are counted using the same logic described form maxT;

  • –interval size in days of overlapping time windows, defaults to 10;

  • –minCoF minimum frequency for high frequency genomic variants, defaults to 0.01;

  • –minP minimum persistence (number of overlapping time windows) for high freq genomic variants: only genomic variants that have a high frequency (>=minCoF) in at least this number of distinct time windows will be included in subsequent analyses, defaults to 3;

  • –outdir output directory. This directory will hold the output files, including lists of high frequency alleles. Defaults to ./metadataAF.

Execution

Please see below for a valid example:

perl computeAF.pl --file linearDataSorted.txt #(where linearDataSorted.txt is is the file with metadata in HaploCoV format)

The output will be stored in the directory specified by –outdir (defaults to ./metadataAF), and will include:

  • frequency matrices for all the countries and macro-geographic areas (suffix _AFOT.txt, ~200 files one per country). These files provide a detailed report of the prevalence of genomic variants of every named country in the HaploCoV formatted file. Normally users do not need to interact with these files, but they might be useful if you need to obtain prevalence data for a specific genomic variant at a specific place;

  • three genomic variants files containing the lists of high frequency genomic variants, showing a frequency above the user defined threshold for more that the timespan set by the user, at global (global_list.txt), macro-areas (area_list.txt) and country (country_list.txt) level. These files can be used along with augmentClusters.pl to identify novel candidate SARS-CoV-2 lineages/designations.

High frequency alleles files from the github repo

Any of global_list.txt, area_list.txt or country_list.txt can be used to provide the list of genomic variants used to expand lineages/sub-lineages of SARS-CoV-2 by augmentClusters.pl. Please see Chiara et al 2022 for a detailed discussion of the implications. A copy of each file is found also in the HaploCoV github repository. These files are updated/regenerated to incorporate new data on a bi-weekly basis (every Wednesday). If you do not want to compute high frequency genomic variants yourself, you can download the files directly from github. On a unix system this can be done by using the wget command . For example:

  1. global_list.txt

wget https://raw.githubusercontent.com/matteo14c/HaploCoV/master/global_list.txt

  1. area_list.txt

wget https://raw.githubusercontent.com/matteo14c/HaploCoV/master/area_list.txt

  1. countries_list.txt

wget https://raw.githubusercontent.com/matteo14c/HaploCoV/master/country_list.txt