2. Installation and Setup - ac812/XWAS-FUSION GitHub Wiki
Welcome to the XWAS-FUSION wiki!
To install XWAS-FUSION:
- Install FUSION by following the Installation instructions.
- Place the XWAS-FUSION scripts in the same directory as the FUSION scripts in step 1.
A reference panel is required in order to generate the imputed summary stats for your data. Below are instructions on the setup required to be able to generate weights for your data.
To generate weights you would need the following:
- Expression files: The expression (or regressions if you have adjusted your data) of each marker in a tab-delimited file with the following columns:
markerID | IndID1 | Ind2 | ... |
---|---|---|---|
cg00000165 | 0.97 | -1.12 | ... |
cg00000363 | -0.43 | 0.78 | ... |
... | ... | ... | ... |
There should be one file for each chromosome and the file names should be expression<i>.txt
where i
is the chromosome number. Each chromosome file must contain the same individuals. Below is an example of this setup:
-
Map files: Map files for the markers present in the
expression<i>.txt
files in (1).
- Single locus marker: If the marker is represented by one position on the genome, e.g. as in SNPs, CpGs, then the map file should have the following format:
markerID | chr | position |
---|---|---|
cg00000165 | 1 | 91194674 |
cg00000363 | 1 | 230560793 |
... | ... | ... |
- Range locus marker: If the marker is represented by a range of loci on the genome, such as a start position and an end position, e.g. as in genes, proteins, then the map file should have the following format:
markerID | chr | posStart | posEnd |
---|---|---|---|
protein1 | 1 | 91194674 | 91197076 |
protein2 | 1 | 230560793 | 230671112 |
... | ... | ... | ... |
One file for each chromosome is required and the file names should be map<i>.txt
where i
is the chromosome number. Below is an example of this setup:
-
PLINK files: Binary PLINK files containing genotype data for the same individuals specified in the
expression<i>.txt
files in (1). One file for each chromosome is required and the file names should begeno<i>.txt
wherei
is the chromosome number. Each chromosome file must contain the same individuals specified in theexpression<i>.txt
files as specified in (i). Below is an example of this setup:
-
Configuration file: A configuration containing the settings to run XWAS-FUSION. The format should be as the
example.config
file below. The following fields are required in the configuration file:
workingDir
: Path to the working directory (where you will be saving the data generated by the scripts).
queue
: name of the queue for your SGE scheduler. (This is necessary to parallelise processes on your cluster).
markerPositionSingle
: TRUE
if the marker is represented by one position, FALSE if the marker is represented by a range (start and end) position.
expressionDir
: Path to the expression<i>.txt
files (as specified in step 1).
mapDir
: Path to the map<i>.txt
files (as specified in step 2).
genoDir
: Path to the chromsome PLINK files (as specified in step 3).
ldrefDir
: Path to the decompressed LDREF directory.
pathPLINK
: Path to the PLINK executable.
pathGemma
: Path the Gemma executable.
pathGCTA
: Path to the gcta executable.
path_qsub
: Path to the qsub executable (for the SGE grid - use which).
pathRLIBS
: R_LIBS path (directory/ies of your R libraries).
email
: email address to send an email to when the XWAS-FUSION finished computation.
workingDir : /home/Alexia/XWAS-Fusion/Example
queue: all.q
markerPostionSingle: TRUE
expressionDir: /home/Alexia/XWAS-Fusion/Data/expressions
mapDir : /home/Alexia/XWAS-Fusion/Data/map
genoDir : /home/Alexia/XWAS-Fusion/Data/plink
ldrefDir : /home/Alexia/XWAS-Fusion/Data/LDREF
pathPLINK : /bin/plink2
pathGemma : /bin/gemma
pathGCTA : /bin/gcta_nr_robust
path_qsub: /bin/qsub
pathRLIBS: /home/Alexia/R/3.3:/bin/R/
email: [email protected]