kmds(1) control for population structure


Control for population structure. Filter kmers and create a matrix representing population structure.

This program belongs to seer(1) (Sequence Element (kmer) Enrichment Analysis).


1) filter and subsample with --no_mds and --size

2) combine, and do metric multidimensional scaling with --combine_mds

Required options:

-k [ --kmers ] arg
dsm kmer output file
-p [ --pheno ] arg
.pheno metadata

MDS options:

-o [ --output ] arg
output prefix for new dsm file
do not perform MDS; output subsampled matrix instead
write csv of distance matrix
--mds_concat arg
list of subsampled matrices to use in MDS. Performs only MDS; implies --no_filtering
--pc arg (=3)
number of principal coordinates to output
--size arg (=1000000)
number of kmers to use in MDS
--threads arg (=1)
number of threads. Suggested: 4

Filtering options:

turn off all filtering and do not output new kmer file
--max_length arg (=100)
maximum kmer length
--maf arg (=0.01)
minimum kmer frequency
--min_words arg
minimum kmer occurrences. Overrides --maf
--chisq arg (=10e-5)
p-value threshold for initial chi squared test. Set to 1 to show all

Other options:

-h [ --help ]
full help message


Filter kmers and create a matrix representing population structure with kmds
kmds -k dsm_input.txt.gz --pheno metadata.pheno -o filtered

To spread this process out, run the following command on each dsm file

kmds -k dsm_input.txt.gz --pheno metadata.pheno --no_mds --size 10000


This manpage was written by Andreas Tille for the Debian distribution and can be used for any other usage of the program.