Microsatellite markers are short, highly variable, multi-repeat DNA
sequences (aka short tandem repeats) that appear throughout the genome
and can be used to estimate population genetic metrics (Silva, Liu, and Blanton 2006; Vieira et al.
2016). These markers are frequently evaluated using fragment
analysis which is based on Sanger sequencing. The
pooledpeaks R package provides tools to analyze fragment
analysis results (.fsa files). It provides functions that fall in three
subcategories: 1) peak scoring, 2) data manipulation, and 3) genetic
analysis. The package was designed for the use of microsatellite markers
on pooled parasite samples, but the peak scoring functions are
applicable to any fragment analysis. The peak scoring functions were
partially adapted from Fragman, a package designed to score
microsatellite markers in cranberries (Covarrubias-Pazaran et al. 2016). Although
Fragman works for the older file version, newer versions cannot be read.
In addition to revising this outdated function, we also added features
including expanded scoring parameter options and exporting resulting
scoring plots as a pdf file for review. The data manipulation functions
were created to clean and format the data from the called peaks and
transform them into allele frequencies. These frequencies can then be
input into the genetic analysis functions for calculation of diversity
and differentiation measures adapted from a range of papers (Long et al. 2022; Jost 2008; Nei 1973; Foulley and
Ollivier 2006; Chao et al. 2008). An in-depth walk-through of how
to use the analysis pipeline can be found in the vignette.
While a plethora of methods exist for downstream statistical analysis of allele frequencies, processing raw fragment data is limited by available software. Of the limited software that can read the .fsa binary raw data file format, nearly all require purchase or registration, are primarily built for Windows, are inefficient for analyzing large batches of files, and are highly dependent on individual researcher experience. Additionally, a previous R package allowing for the analysis of .fsa files is incompatible with the updated file version. When using fragment analysis for microsatellite markers on pooled samples, once the raw data is extracted and scored, it must be cleaned and transformed into allele frequencies using a second software, such as Excel, which is limited in its capacity for automation and version control. Another platform shift is often required to analyze the resulting allele frequencies. These factors highlight the need for a comprehensive scoring and analysis pipeline that is open-source, offline, reproducible, consistent between researchers, and that does not require platform switching between steps.
This package is currently being used to analyze genetic clustering of Schistosoma mansoni pooled egg samples from four Brazilian communities, as well as the relatedness of Schistosoma haematobium populations around Lac de Guiers in Senegal and from Gabon.
This work was financially supported by the NIH as part of 1R01AI121330.