runFeatureSelection {netDx} | R Documentation |
Run GeneMANIA cross-validation with a provided subset of networks
runFeatureSelection(trainID_pred, outDir, dbPath, numTrainSamps = NULL, incNets = "all", orgName = "predictor", fileSfx = "CV", verbose = FALSE, numCores = 2L, JavaMemory = 6L, verbose_runQuery = FALSE, debugMode = FALSE, ...)
trainID_pred |
(char) vector with universe of predictor class patients (ie all that can possibly be included in the query file |
outDir |
(char) directory to store query file and GM results |
dbPath |
(char) path to GeneMANIA generic database with training population |
numTrainSamps |
(integer) number of training samples in total leave blank to use 5 training samples in order to save memory |
incNets |
(char) vector of networks to include in this analysis (features/pathway names). Useful for subset-based feature selection |
orgName |
(char) organism name for GeneMANIA generic database. The default value will likely never need to be changed. |
fileSfx |
(char) file suffix |
verbose |
(logical) print messages |
numCores |
(logical) num parallel threads for cross-validation |
JavaMemory |
(integer) memory for GeneMANIA run, in Gb. |
verbose_runQuery |
(logical) print messages for runQuery() |
debugMode |
(logical) when TRUE runs jobs in serial instead of parallel and prints verbose messages. Also prints system Java calls and prints all standard out and error output associated with these calls. |
... |
args for |
Creates query files, runs GM for 10-fold cross validation.
No value. Side effect of generating feature scores.
data(MB.pheno) dbPath <- system.file("extdata","dbPath",package="netDx") runFeatureSelection(MB.pheno$ID[which(MB.pheno$STATUS%in% 'WNT')], tempdir(),dbPath,103L)