Takes a qlist and combines several repeats for each K into a single file along with a parameter file suitable for input to CLUMPP. The two output files are organised into folders by K. This function only creates the input files for CLUMPP. It does not run CLUMPP.
clumppExport( qlist = NULL, prefix = NA, parammode = NA, paramrep = NA, exportpath = NULL, path = NULL )
| qlist | A qlist (list of dataframes). An output from |
|---|---|
| prefix | A character prefix for folder names. By default, set to 'pop'. |
| parammode | A numeric 1, 2 or 3 indicating the algorithm option for CLUMPP paramfile. Calculated automatically by default. Set this value to 3 if CLUMPP runs too long. See details. |
| paramrep | A numeric indicating the number of repeats for CLUMPP paramfile. Calculated automatically by default. See details. |
| exportpath | The path to export output files. To use current working directory, set |
| path | Deprecated. Use exportpath. |
The combined file and paramfile are written into respective folders named by K.
This function only generates the files needed to run CLUMPP.
The CLUMPP executable can be downloaded and run for downstream steps. It can
be obtained from https://web.stanford.edu/group/rosenberglab/clumpp.html.
Please remember to cite CLUMPP if you use it.
When multiple repeats are run for each K in runs, the order of
clusters may be jumbled for each run. Therefore, when plotting multiple runs
within each K, the colours cannot be assigned correctly. The software CLUMPP
helps to overcome this issue by reordering the clusters correctly. This
function clumppExport() takes multiple runs for each K and combines
them into a single file and generates a parameter file for easy use with
CLUMPP. Further details for CLUMPP can be found here: Jakobsson, M., and
Rosenberg, N. A. (2007). CLUMPP: a cluster matching and permutation program
for dealing with label switching and multimodality in analysis of population
structure. Bioinformatics, 23(14), 1801-1806.
parammode
The parammode (M) is the type of algorithm used. Option 1 is 'FullSearch'
(takes the longest time), option 2 is 'Greedy' and option 3 is 'LargeKGreedy'
(fastest). If clumpp takes more than a few minutes, consider changing
parammode to a higher number (ex. from 2 to 3), or open the exported
paramfile and manually change GREEDY_OPTION to 3.
The parammode and paramrep for CLUMPP paramfile is set based on this
calculation.
X <- factorial(k)*((runs*(runs-1))/2)*k*ind, where k is number of
populations, runs is number of runs for k and ind is number of individuals.
If X <= 100000000, then parammode is 2 and paramrep is 20, otherwise
parammode is 3 and paramrep is set to 500.
To find out more about parammode (algorithm type) and paramrep (repeats),
refer to CLUMPP documentation.
See the vignette for more details.
if (FALSE) { # generate input files for CLUMPP from STRUCTURE files sfiles <- list.files(path=system.file("files/structure",package="pophelper"), full.names=TRUE) clumppExport(readQ(sfiles),exportpath=getwd()) # generate input files for CLUMPP from ADMIXTURE files afiles <- list.files(path=system.file("files/admixture",package="pophelper"), full.names=TRUE) clumppExport(readQ(afiles),exportpath=getwd()) }