GRMA tutorial:

You can use the provided requirements.txt file (run "pip install -r requirements.txt") to have the
needed versions of packages installed on your system.  If this conflicts with packages or their
needed versions already on your system, you could try using a virtual environment to run GRMA.
See virtualenv (https://virtualenv.pypa.io/en/latest/) or conda
(https://anaconda.org/channels/anaconda/packages/conda/overview), for example, or use whichever
other similar tool you like best.

See below for full usage (or run "grma.py -h"), but for the purposes of a simple demonstration of GRMA and verifying
that it works on your system, run:

grma.py --bfile tutorial/tutorial --rel-pedigree tutorial/tutorial.kin --rel-thresh FS --out tutorial/user_FS

If grma.py is not found, try using "./grma.py" or make sure the current working directory is added to your PATH.

This should produce files user_FS.res and user_FS.log in the tutorial directory.  The log file will show
some status and processing steps that GRMA went through during the analysis.  For a more extended log, you can
add the --verbose option.  The .res file has the actual GRMA results, in this case using the full sibling
relationship threshold.

Next, try running:

grma.py --bfile tutorial/tutorial --rel-pedigree tutorial/tutorial.kin --rel-thresh 3 --out tutorial/user_3

This should produce files user_3.res and user_3.log in the tutorial directory.
In this case, the .res file has results using the degree 3 relationship threshold.


Compare the "user" files with the provided "tutorial" files to make sure things are running correctly.
The results should be very close (different package versions and hardware could introduce negligible deviations).

You can also run "pytest" in the main GRMA folder to run through the suite of GRMA tests (especially
useful if the results of the previous steps differ too much from the provided results, since it could
provide information about what portions of the code might be running into issues)


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

Usage info:

usage: ./grma.py [-h] --bfile GLOB_PATH [GLOB_PATH ...] [--fam FILE] [--pheno FILE] [--covar FILE] --rel-pedigree FILE [--rel-thresh THRESHOLD] [--id-list FILE] [--snp-list FILE]
                 [--out FILE_PREFIX] [--snps-per-block SNPS_PER_BLOCK] [--quiet | --verbose]

options:
  -h, --help            show this help message and exit

Input Specifications:
  --bfile GLOB_PATH [GLOB_PATH ...]
                        Paths to PLINK 1 binary files. See python glob module for documentation on the string(s) to be provided here (full path with support for "*", "?", and "[]").
                        These strings should be encased in quotes.
  --fam FILE            Optional input to specify full path and filename of input fam file. Overrides --bfile flag if specified
  --pheno FILE          Optional input to specify a (Plink-style) phenotype file: https://www.cog-genomics.org/plink/1.9/input#pheno
  --covar FILE          Optional input to specify a (Plink-style) covariates file: https://www.cog-genomics.org/plink/2.0/input#covar
  --rel-pedigree FILE   Path to KING-formatted relatedness file (ASCII). Needs the following columns: ['FID1', 'ID1', 'FID2', 'ID2', 'Kinship', 'InfType']

Analysis Options:
  --rel-thresh THRESHOLD
                        Relatedness threshold. Can be one of ['FS', '1', '2', '3']. Default is 1

Input Filtering Options:
  --id-list FILE        Optional input to specify a whitespace-delimited sample ID file of sample FIDs and IIDs to include. No header allowed. See Plink flag --keep.
  --snp-list FILE       Optional input to specify a whitespace-delimited variant ID file of variant IDs to include. No header allowed. See Plink flag --extract.

Output Specifications:
  --out FILE_PREFIX     Full prefix of output files (logs, sumstats results, etc.). If not set, [current working directory]/grma will
                        be used. Note: The containing directory specified must already exist.

General Options:
  --snps-per-block SNPS_PER_BLOCK
                        Number of SNPs to process at a time. Default is 100
  --quiet               This option will cause the program to limit logging and terminal output to warnings and errors and reduce output compared to the default/standard logging mode. It
                        is mutually exclusive with the --verbose option.
  --verbose             This option will greatly increase the logging and terminal output of the program compared to the default/standard logging mode. This is useful for debugging and
                        greater visibility into the processing that is occurring. It is mutually exclusive with the --quiet option.