tni.permutation: Inference of transcriptional networks.

Description Usage Arguments Value Author(s) References See Also Examples

Description

This function takes a TNI object and returns a transcriptional network inferred by mutual information (with multiple hypothesis testing corrections).

Usage

1
2
3
tni.permutation(object, pValueCutoff=0.01, pAdjustMethod="BH", globalAdjustment=TRUE,
        estimator="spearman", nPermutations=1000, pooledNullDistribution=TRUE, 
        boxcox=TRUE, parChunks=NULL, verbose=TRUE)

Arguments

object

a preprocessed object of class 'TNI' TNI-class.

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 powerTransform and bcPower.

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)

Value

a mutual information matrix in the slot "results" containing a reference transcriptional network, see 'tn.ref' option in tni.get.

Author(s)

Mauro Castro

References

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.

See Also

TNI-class

Examples

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
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)

RTN documentation built on Nov. 12, 2020, 2:02 a.m.