rSpagedi

rSpagedi is a package aimed at interfacing the genetics program SPAGeDi (Spatial Pattern Analysis of Genetic Diversity) with R. The current version of rSpagedi focuses on running analyses of fine-scale spatial genetic structure (FS-SGS).

To find out more about SPAGeDi, please visit the website.

The main components of rSpagedi are:

1. Running SPAGeDi analyses in R via the terminal
2. Reading the output of SPAGeDi analyses into R
3. Calculating the popular Sp statistic, which estimates the strength of FS-SGS
4. Creating spatial autocorrelation plots

September 2019 update

Over the years, I've had people reach out to me with questions about running rSpagedi on their own datasets. While I do think rSpagedi could be useful for some people, its functionality is pretty limited, and I do not foresee having the ability expand on this in the future. My current advice is that if you are looking for a way to run SPAGeDi in a programmatic way (e.g., run analyses on many datasets or across many parameter values), I would suggest looking into how to do this via the command line and Bash scripting. Here is a short tutorial on bash scripting that may be helpful. Ultimately, this will be a much more flexible and powerful approach than using the rSPagedi package, and it is the approach that I would take if I could go back in time.

The SPAGeDi manual has a short section on running it from the command line with some advice:

"The program can also be launched using a command line or via another application, which can be useful to analyse numerous data sets obtained for instance by simulations. A command file that contains the keystrokes used to run an analysis can be associated to automatize the runs (e.g. “spagedi < cmds.txt” where cmds.txt is the file with the keystrokes commands). On Unix-derived systems, the program “tee” can be used to record keystrokes for playback later (e.g. “tee cmds.txt | spagedi” to record and “spagedi < cmds.txt” to repeat)."

===============

rSpagedi is no longer in development, and certainly contains many bugs. If you have any questions, please get in touch via github or email - [email protected]

===============

To install rSpagedi, you must first make sure the package 'devtools' is installed. This will allow you to install rSpagedi directly from github.

install.packages("devtools")

devtools::install_github("lukembrowne/rSpagedi")

=================

Here's an example of a typical workflow...

• use runSpagedi() to run Spagedi through R
  • Note that SPAGeDi must be installed on your system!
  • It is also configured to run the kinship coefficient of Loiselle et al. 1995
• use makeSpagediList() to read back into R in the data file created by the Spagedi run
• use SpSummary() to calculate Sp statistics
• use plotAutoCor() to make an autorcorrelation plot

library(rSpagedi)

## Run SPAGeDi through R console with various options for analysis
# This reads in the SPAGeDi formatted data file located in your working directory, and
# then writes the SPAGeDi output file to your working directory
runSpagedi(input_name = "example_input_file.txt",
           output_name = "output_file.txt", 
           categories_present = TRUE, 
           perm = TRUE, n_perm = 100,
           rest_reg = TRUE, max_regr_dist = 160,
           jackknife = TRUE, 
           ar = FALSE, min_ar = 100)

## Read back in SPAGeDi into R
spagedi_output <- makeSpagediList(path_to_out = "output_file.txt")

## Calculate Sp Statistics
SpSummary(spagedi_output)
#> 
#> 
#> Mean Sp across all loci       ==  0.101427 
#> -------
#> Mean Sp across jacknifed loci ==  0.1011885 
#> -------
#> Sp by loci
#>       Ob03       Ob10       Ob19        Ob4       Ob16       Ob12 
#> 0.10549631 0.11325540 0.11171602 0.03014093 0.08078488 0.12538566 
#>       Ob22       Ob06       Ob07       Ob23       Ob11 
#> 0.10003766 0.13839420 0.13311573 0.15410320 0.04693298

## Create spatial autocorrelation plot
plotAutoCor(spagedi_output, max_dist = 200)

# You can also retrive the raw data produced by SPAGeDi like this..

spagedi_output$perm # Info about permutation tests
spagedi_output$dist # Info on distance classes
spagedi_output$kin  # Info on kinship estimates
spagedi_output$diversity # Info on diversity estimates

=================