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 latest version from GitHub:

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

Compilation install error

When installing on a Mac or LINUX system, you may get errors when installing during the compilation phase that look like:

─  installing *source* package ‘strataG’ ...
   ** using staged installation
   ** libs

...

   ld: warning: search path '/opt/gfortran/lib/gcc/aarch64-apple-darwin20.0/12.2.0' not found
   ld: warning: search path '/opt/gfortran/lib' not found
   ld: library 'gfortran' not found
   clang++: error: linker command failed with exit code 1 (use -v to see invocation)
   make: *** [strataG.so] Error 1
   ERROR: compilation failed for package ‘strataG’
─  removing ‘/private/var/folders/rx/z5h877kx2rx_85q95ct3fdfr0000gn/T/RtmpgBBQQU/Rinstfc0912a21e41/strataG’
         -----------------------------------
   ERROR: package installation failed
Error: Failed to install 'strataG' from GitHub:
  ! System command 'R' failed

If you see this, follow these steps to update the compilation-related paths in your Makeconf file:

1. Confirm that you have gfortran and gcc installed. Open Terminal and type

<me> ~ % which gfortran
/usr/local/bin/gfortran
<me> ~ % which gcc
/usr/bin/gcc

If you see "gfortran not found", then you need to install it. First, if you are on a Mac, make sure you have Xcode Command Line Tools installed:

xcode-select --install

If which gfortran still returns gfortran not found and you have homebrew then try:

brew install gcc

2. Locate your Makeconf file. Check /Library/Frameworks/R.framework/Resources/etc:

<me> ~ % ls /Library/Frameworks/R.framework/Resources/etc
Makeconf	Renviron	javaconf	ldpaths		repositories

3. Open Makeconf in a text editor and find the lines labeled #Fortran.

4. Update the FC and F77 paths to match the location given in which gfortran:

FC = /usr/local/gfortran/bin/gfortran -arch arm64
F77 = /usr/local/gfortran/bin/gfortran

5. Search in the same gfortran directory for gcc, perhaps in lib. Update FLIBS to match the paths for the libraries as below:

FLIBS =  -L/usr/local/gfortran/lib/gcc/aarch64-apple-darwin23/14.1.0 -L/usr/local/gfortran/lib -lgfortran -lemutls_w -lquadmath

For more help, also check out this thread on Stack Overflow.

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. 2025. 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

• submit suggestions and bug-reports: https://github.com/ericarcher/strataG/issues
• send a pull request: https://github.com/ericarcher/strataG/
• e-mail: [email protected]

---

Versions

2.5.5 (devel)

• IMPORTANT: There was an error in computing observed hetrozygosity (in heterozygosity, summarizeLoci(), and summarizeInds()) where the number of genotypes missing data were not being taken into account. This would lead to an underestimate in heterozygosity that was directly proportional to the amount of missing data.
• added zygosity()
• added diagnosability()
• added microhaplot2rubias()
• added qual2prob()
• made rfPermute and rmetasim packages Suggested
• 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
• added gtypesRF(), sequenceRF(), and gtypes2rfDF().

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().

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.

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.

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

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.

1.0

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