Skip to content

Latest commit

 

History

History
60 lines (35 loc) · 1.9 KB

LdPruning.md

File metadata and controls

60 lines (35 loc) · 1.9 KB

LdPruning

This command LD-prunes result files.

General considerations

Files need to be formatted as detailed here.

Command line

java -Xmx16G -cp your/folder/triogen-X.Y.Z/triogen-X.Y.Z.jar no.uib.triogen.cmd.ld_pruning.LdPruning [parameters]

Note: you need to replace your/folder by the folder where the release is installed, and Z.Y.Z by the version number.

Standard parameters

-h/--help                 Display help text
-v/--version              Display version

Mandatory Parameters

-res/results              The results file to prune. Can be any text file, gzipped or not.
-l/--ldMatrix             The LD matrix file as generated using the [_LdMatrix_](LdMatrix.md) command. If LD matrix files are computed per chromosome, replace the chromosome name with '{contig}'.
-f/--fam                  The trio identifiers file.
-o/--out                  The file where to write the results.

Additional Parameters

-r/--minR2                The minimal LD r2 to consider two markers in LD. Default: 0.05. Min value: value used when generating the LD matrix file.
-p/--maxP                 The maximal p-value to consider. Default: 1e-6.
-id/--idColName           The name of the variant id column. Default: 'variantId'.
-pn/--pColName            The name of the p-value column. Default: 'h.intercept.p'.
-cn/--contigColName       The name of the contig column. Default: 'contig'.
-phn/--phenoColName       The name of the phenotype column. Ignored if not provided.
-s/--separator            Separator for the columns. Default: '\t'.

Processing

For each phenotype, variants are selected from lowest p-values until reaching maxP excluding variants with an LD R2 >= minR2 with any of the already selected variants.

Output

The entire lines of the variants passing the pruning are written to the output as they were in the input.