tni.permutation {RTN} | R Documentation |
This function takes a TNI object and returns a transcriptional network inferred by mutual information (with multiple hypothesis testing corrections).
tni.permutation(object, pValueCutoff=0.01, pAdjustMethod="BH", globalAdjustment=TRUE, estimator="spearman", nPermutations=1000, pooledNullDistribution=TRUE, boxcox=TRUE, parChunks=NULL, verbose=TRUE)
object |
a preprocessed object of class 'TNI' |
pValueCutoff |
a single numeric value specifying the cutoff for p-values considered significant. |
pAdjustMethod |
a single character value specifying the p-value adjustment method to be used (see 'p.adjust' for details). |
globalAdjustment |
a single logical value specifying to run global p.value adjustments (when globalAdjustment=TRUE) or not (when globalAdjustment=FALSE). |
estimator |
a character string specifying the mutual information estimator. One of "pearson", "kendall", or "spearman" (default). |
nPermutations |
a single integer value specifying the number of permutations for deriving TF-target p-values in the mutual information analysis. If running in parallel, nPermutations should be greater and multiple of parChunks. |
pooledNullDistribution |
a single logical value specifying to run the permutation analysis with pooled regulons (when pooledNullDistribution=TRUE) or not (when pooledNullDistribution=FALSE). |
boxcox |
a single logical value specifying to use Box-Cox procedure to find a transformation of inferred associations that approaches normality (when boxcox=TRUE) or not (when boxcox=FALSE). Dam et al. (2018) have acknowledged that different RNA-seq normalization methods introduce different biases in co-expression analysis, usually towards positive correlation, possibly affected by read-depth differences between samples and the large abundance of 0 values present in RNA-seq-derived expression matrices. In order to correct this positive correlation bias we suggest using this box-cox correction strategy. See |
parChunks |
an optional single integer value specifying the number of permutation chunks to be used in the parallel analysis (effective only for "pooledNullDistribution = TRUE"). |
verbose |
a single logical value specifying to display detailed messages (when verbose=TRUE) or not (when verbose=FALSE) |
a mutual information matrix in the slot "results" containing a reference transcriptional network,
see 'tn.ref' option in tni.get
.
Mauro Castro
Dam et al. Gene co-expression analysis for functional classification and gene-disease predictions. Brief Bioinform. 2018 Jul 20;19(4):575-592. doi: 10.1093/bib/bbw139.
data(tniData) ## Not run: # preprocessing rtni <- tni.constructor(expData=tniData$expData, regulatoryElements=c("PTTG1","E2F2","FOXM1","E2F3","RUNX2"), rowAnnotation=tniData$rowAnnotation) # linear version (set nPermutations >= 1000) rtni <- tni.permutation(rtni, nPermutations = 100) ## parallel version with SNOW package! #library(snow) #options(cluster=snow::makeCluster(3, "SOCK")) #rtni<-tni.permutation(rtni) #stopCluster(getOption("cluster")) ## End(Not run)