This directory contains domainXplorer, an algorithm to identify correlations between
phenotypes and mutations in protein regions, including domains and disordered regions.

In order to run domainXplorer you will need R and Perl as well as the following 
tab-separated files:

1 - File with mutations, it should have the following columns:
	
	a) Name of the sample
	b) Name of the protein
	c) Position of the mutation
	
2 - File with the phenotype annotations, it should have the following columns:
	
	a) Name of the sample (it should be the same as in the file with mutation data)
	b) Batch (optional, just add a dummy here if you do not want to take this into
	account)
	c) Phenotype. Ideally this should be a continuous variable that quantifies some
	phenotype. We provide an example file with ESTIMATE immune scores

3 - File with the domain annotations with this format:

	a) Name of the protein (it should be the same as in the file with mutation data)
	b..n) Protein regions annotated as NNNN-S-E, where NNNN is the name of the region,
	S is the start position of the region and E is the end position of the region. For
	example if there is a kinase domain (PF00069) between positions 7 and 246 it should
	be in the file as PF00069-7-246

You will also need a directory to store the files and root name for the output.

For example, to reproduce the domainXplorer analysis of ESTIMATE immune annotations of 
TCGA samples in a directory called "immmune_domains" with the root name "tumor_immunology"
use this command:

perl domainXplorer immune_domains/ tumor_immunology missense_mutations_TCGA.txt ESTIMATE_scores.txt protein_regions.txt 0

You can find all the necessary files in the folder "example_files". There are three types
of protein regions in the "protein_regions.txt" file: PFAM domains, annotated with
PFAMscan, disordered regions, annotated with Foldindex, and putative new protein domains,
identified with the AIDA server from Adam Godzik's laboratory.