gmt music smg - Identify significantly mutated genes.
This document describes gmt music smg version 0.04 (2016-01-01 at 17:58:00)
gmt music smg --gene-mr-file=? --output-file=? [--max-fdr=?]
[--skip-low-mr-genes] [--bmr-modifier-file=?] [--processors=?]
... music smg \
--gene-mr-file output_dir/gene_mrs \
(A "gene-mr-file" can be generated using the tool "music bmr
- gene-mr-file Text
- File with per-gene mutation rates (Created using
"music bmr calc-bmr")
- output-file Text
- Output file that will list significantly mutated genes and
- max-fdr Number
- The maximum allowed false discovery rate for a gene to be
considered an SMG
Default value '0.2' if not specified
- skip-low-mr-genes Boolean
- Skip testing genes with MRs lower than the background MR
Default value 'true' if not specified
- noskip-low-mr-genes Boolean
- Make skip-low-mr-genes 'false'
- bmr-modifier-file Text
- Tab delimited multipliers per gene that modify BMR before
testing [gene_name bmr_modifier]
- processors Integer
- Number of processors to use (requires 'foreach' and 'doMC'
Default value '1' if not specified
This script runs R-based statistical tools to identify Significantly Mutated
Genes (SMGs), when given per-gene mutation rates categorized by mutation type,
and the overall background mutation rates (BMRs) for each of those categories
(gene_mr_file, created using "music bmr calc-bmr").
P-values and false discovery rates (FDRs) for each gene in gene_mr_file is
calculated using three tests: Fisher's Combined P-value test (FCPT),
Likelihood Ratio test (LRT), and the Convolution test (CT). For a gene, if its
FDR for at least 2 of these tests is <= max_fdr, it will be output as an
SMG. Another output file with prefix "_detailed" will have p-values
and FDRs for all genes.
- The user can provide a BMR modifier for each gene in the
ROI file, which is a multiplier for the categorized background mutation
rates, before testing them against the gene's categorized mutation rates.
Such a file can be used to correct for regional or systematic bias in
mutation rates across the genome that may be correlated to CpG deamination
or DNA repair processes like transcription-coupled repair or mismatch
repair. Mutation rates have also been associated with DNA replication
timing, where higher mutation rates are seen in late replicating regions.
Note that the same per-gene multiplier is used on each mutation category of
BMR. Any genes from the ROI file that are not in the BMR modifier file will
be tested against unmodified overall BMRs per mutation category. BMR
modifiers of <=0 are not permitted, because that's just silly.
- Genes with consistently lower MRs than the BMRs across
mutation categories, may show up in the results as an SMG (by CT or LRT). If
such genes are not of interest, they may be assigned a p-value of 1. This
should also speed things up. Genes with higher Indel or Truncation rates
than the background will not be skipped even if the gene's overall MR is
lower than the BMR. If bmr-modifiers are applied, this step uses the
modified BMRs instead.
Qunyuan Zhang, Ph.D.
Cyriac Kandoth, Ph.D.
Nathan D. Dees, Ph.D.