Skip to content

strataG is a toolkit for haploid sequence and multilocus genetic data summaries, and analyses of population structure.

Notifications You must be signed in to change notification settings

EricArcher/strataG

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

CRAN version CRAN last day downloads CRAN last week downloads CRAN last month downloads CRAN total downloads
Zenodo DOI
Travis-CI Build Status AppVeyor Build Status

strataG

Description

strataG is a toolkit for haploid sequence and multilocus genetic data summaries, and analyses of population structure. One can select select specific individuals, loci, or strata using standard R '[' indexing methods. . The package contains functions for summarizing haploid and diploid loci (e.g., allelic richness, heterozygosity, haplotypic diversity, etc.), and haploid sequences by locus and by strata as well as functions for computing by-site base frequencies and identifying variable and fixed sites among strata. There are both overall and pairwise standard tests of population structure like PHIst, Fst, Gst, and Jost's D. If individuals are stratified according to multiple schemes, these stratifications can be changed with the stratify() function and summaries or tests can be re-run on the new object. The package also includes wrappers for several external programs like fastsimcoal2, STRUCTURE, and mafft. There are also multiple conversion functions for data objects for other population packages such as adegenet, pegas, and phangorn.

Installation

To install the stable version using install.packages requires an extra repo to be made available to the install.packages function prior to install as the strataG is not available via CRAN:

options(repos = c(
            zkamvar = 'https://zkamvar.r-universe.dev',
            CRAN = 'https://cloud.r-project.org'))

install.packages('strataG')

NB! Make sure that you have installed the development version of the dependency sprex prior to installing strataG

To install the latest version from GitHub including the development version of sprex:

# make sure you have Rtools installed
if (!require('devtools')) install.packages('devtools')
# install sprex development version
devtools::install_github("ericarcher/sprex")
# install strataG latest version
devtools::install_github('ericarcher/strataG', build_vignettes = TRUE)

Vignettes

Vignettes are available on several topics:

  • Creating and manipulating gtypes ("gtypes")
  • Genotype and sequence summaries ("summaries")
  • Working with sequences ("sequences")
  • Tests of population structure ("population.structure")
  • Installing external programs ("external.programs")

To see the list of all available vignettes:

browseVignettes("strataG")

To open a specific vignette:

vignette("gtypes", "strataG")

There is also a tutorial detailing running fastsimcoal2 through strataG available through the function fscTutorial().

Citation

The paper can be obtained here, and is cited as (preferred):

Archer, F. I., Adams, P. E. and Schneiders, B. B. (2016), strataG: An R package for manipulating, summarizing and analysing population genetic data. Mol Ecol Resour. doi:10.1111/1755-0998.12559

If desired, the current release version of the package can be cited as:

Archer, F. 2016. strataG: An R package for manipulating, summarizing and analysing population genetic data. R package version 1.0.6. Zenodo. http://doi.org/10.5281/zenodo.60416

Contact

version 2.5.01 (devel)

  • removed melt from structurePlot
  • fixed ldNe error when one individual is present
  • fixed mafft error and now have mafft .fasta files written to temporary file rather than working directory
  • fixed error with readGenData() not recognizing NAs.
  • fixed error with fs2gtypes() not formatting multi-block DNA sequence data as gtypes properly

version 2.4.905

  • Deleted functions: alleleFreqFormat, as.array.gtypes
  • Changed structure of gtypes object, making it no longer compatible with previous versions
  • Fixed and enhanced arlequinRead() so that it will read and parse all .arp files. Added arp2gtypes() to create gtypes object from parsed .arp files.
  • Improved performance of several standard summary functions, most notably dupGenotypes().
  • Full rework of fastsimcoal2 wrapper.
  • Removed strataGUI().

version 2.1

  • fixed error in ldNe when missing data are present
  • added STANDARD marker type to fastsimcoal
  • added na.rm = TRUE to calculation of mean locus summaries by strata in summary.gtypes. This avoids NaNs when there is a locus with genotypes missing for all samples.
  • explicitly convert x to a data.frame in df2gtypes in case it is a data.table or tibble.

version 2.0.2

  • NOTE: In order to speed up indexing the data in large data sets, this version changes the underlying structure of the gtypes object by replacing the @loci data.frame slot with a @data data.table slot. The data.table has a id character column, a strata character column, and every column afterwards represents one locus. The @strata slot has been removed.
  • The loci accessor has been removed.
  • Added as.array which returns a 3-dimensional array with dimensions of [id, locus, allele].
  • The print (show) function for gtypes objects no longer shows a by-locus summary. The display was getting too slow for data sets with a large number of loci.
  • The summary function now includes by-sample results.
  • Fixed computational errors in population structure metrics due to incorrect sorting of stratification.
  • Added maf to return minimum allele frequency for each locus.
  • Added ldNe to calculate Ne.
  • Added expandHaplotypes to expand the haplotypes in a gtypes object to one sequence per individual.

version 1.0.6

  • Added read.arlequin back. Fixed missing function error with write.arlequin.
  • Added summarizeSamples
  • Changed evanno from base graphics to ggplot2
  • Updated logic in labelHaplotypes to assign haplotypes if possible alternative site combinations match a present haplotype
  • Added Zenodo DOI
  • Added shiny app (strataGUI) for creating gtypes objects, QA/QC, and population structure analyses
  • Added type argument to structurePlot to select between area and bar charts
  • Changed haplotypeLikelihoods to sequenceLikelihoods
  • neiDa now creates haplotypes before calculating metric
  • Fixed error in writePhase that was creating improper input files for PHASE

version 1.0.5

  • Fixed error in dupGenotypes, propSharedLoci, and propSharedIDs where missing genotypes were not being properly counted.
  • Added as.data.frame.gtypes.
  • Removed gtypes2df.
  • Added arguments to as.matrix.gtypes to include id and strata columns in output.
  • Removed the jmodeltest function as this functionality is available in the modeltest function in the phangorn package.
  • Added conversion functions gtypes2phyDat and phyDat2gtypes to facilitate interoperability with the phangorn package.
  • Removed read.arlequin.
  • Added alleleNames accessor for gtypes object, which returns list of allele names for each locus.

version 1.0

  • New version with different gtypes format from previous versions. See vignettes for instructions and examples.