Quick Start Guide

PRIMUS-2.8.pl is a stand alone perl script that can be run on the commandline. For documentation and usage type:

perl PRIMUS-2.8.pl -help

or read the PRIMUS_documentation file.

RUNNING EXAMPLE DATA

Inside the example_data directory there are four files:

1. example.genome - contain the pairwise IBD estimates for the sample data. This is the file format of PLINK's IBD estimate output.
2. example.ped - standard ped file format; however, the parent ids have been left as 0 (they are not needed).
3. example.btrait - this file contains a sample id column and a binary trait column (numbered 1 and 2). The binary trait column matches the 6th colum of the ped file.
4. example.qtrait - contains sample id column and a quantitative trait value in the second column.

Here are the commands to run the example data.

1. Run PRIMUS on to get the maximum unrelated set from the example.genome file from the PRIMUS/bin directory

perl PRIMUS-2.8.pl -input ../example_data/example.genome

There should be standard output to the terminal window that looks like this:

Trait file: size
Relatedness_file: ../example_data/example.genome
Threshold: 0.1
Selection criteria are based on the following:
size
Writing results files to ../example_data/example.genome_results/
Loading data...
Loading trait data: size
done.
colapsing networks...
done.
Checking for large networks...
done.
writing out independent set...
done.

and results files should have been written to "PRIMUS/example_data/example.genome_results/". There should be 3 files in this directory: example.genome_maximum_independent_set, example.genome_network7.dot, and example.genome_networks
The maximum set of unrelated or independent individuals is in example.genome_maximum_independent_set, and it should have samples 56, 57, 58, and 61. This is the largest set of unrelated individuals in this pedigree. This set contains 3 unaffected individuals and 1 affected individual.

1. Run PRIMUS on to get the a set of unrelated individuals while maximizing the number of affected individuals

perl PRIMUS-2.8.pl -input ../example_data/example.genome -ped ../example_data/example.ped

You should see that the selection criteria has changed from

Selection criteria are based on the following:
size

and is now

Selection criteria are based on the following:
../example_data/example.ped
size

You should also see that the file example.genome_maximum_independent_set is different. It should contain sample 61, 70, and 58 (2 affecteds and 1 unaffected).
This weighting criteria sacrificed the maximum size of the unrelated set to maximize the number of affected individuals.

1. Run PRIMUS to first select for maximum size and then to weight for affected individuals

perl PRIMUS-2.8.pl -input ../example_data/example.genome -trait size -ped ../example_data/example.ped

By putting the "-trait size" option first, it will first weight on size and then the ped file's affected status column.
Notice the selection criteria change

Selection criteria are based on the following:
size
../example_data/example.ped

This will produce the same results as example run 1.

1. Run PRIMUS to weight on a binary trait other than the ped file

perl PRIMUS-2.8.pl -input ../example_data/example.genome -low/btrait ../example_data/example.btrait

will select an unrelated set of idividuals that maximizes the number of individuals with a btrait = 2.

or

perl PRIMUS-2.8.pl -input ../example_data/example.genome -high/btrait ../example_data/example.btait

will select an unrelated set of idividuals that maximizes the number of individuals with a btrait = 1.

1. Run PRIMUS weighitng for a quantitative trait

perl PRIMUS-2.8.pl -input ../example_data/example.genome -low/qtrait ../example_data/example.qtrait

The quantitative weighting method is only able to weight as a secondary weighting criteria, meaning that you must weight by size or a binary trait first.
If the only weighting trait provided is a qtrait (like in the example above), the program will weight for size first, and then the qtrait.
For this particular example, weighting by a qtrait affect the results because there is a single maximum set, and that set will be selected regarless of weighting.
However, in family networks where there are multiple maximum sets, it will select the one with the lowest average qtrait values.
You can also do the following command if you wish to weight differently using the qtrait:

perl PRIMUS-2.8.pl -input ../example_data/example.genome -high/qtrait ../example_data/example.qtrait
perl PRIMUS-2.8.pl -input ../example_data/example.genome -tails/qtrait ../example_data/example.qtrait
perl PRIMUS-2.8.pl -input ../example_data/example.genome -mean/qtrait ../example_data/example.qtrait
perl PRIMUS-2.8.pl -input ../example_data/example.genome -5/qtrait ../example_data/example.qtrait

1. Weighting on a binary trait and then a quantitative trait

perl PRIMUS-2.8.pl -input ../example_data/example.genome -high/btrait ../example_data/example.btrait -low/qtrait ../example_data/example.qtrait