DCARS

the DCARS function

DCARS(dat, xname, yname, W = NULL, rangeMin = 0, wcormin = 0,
  statmin = 0, extractTestStatisticOnly = FALSE,
  extractWcorSequenceOnly = FALSE, plot = FALSE, niter = 100,
  extractPermutationTestStatistics = FALSE, verbose = FALSE, ...)

Arguments

dat	a genes x samples gene expression rank matrix, should be already converted to ranks with first column lowest survival and last column highest survival
xname	name of row of dat to test together with yname
yname	name of row of dat to test together with xname
W	weight matrix for weighted correlations,
rangeMin	minimum range of weighted correlation vector to include for permutation testing
wcormin	minimum absolute value weighted correlation vector to include for permutation testing
statmin	minimum value DCARS test statistic to include for permutation testing
extractTestStatisticOnly	if TRUE, extract only the DCARS test statistic without permutation testing
plot	if TRUE plot observed weighted correlatin vector
niter	number of iterations for permutation testing
extractPermutationTestStatistics	if TRUE, extract only the DCARS test statistic without permutation testing
verbose	if TRUE, print updates
...	additional arguments passing on to weightMatrix()
extractWcorSequence	if TRUE, extract only the weighted correlation vector without permutation testing

Value

either single value (p-value or test statistic), vector (local weighted correlation), or list (combination of above) depending on the input parameters

Examples

data(STRING)
data(SKCM)
SKCM_rank = t(apply(SKCM,1,rank))

# highly significantly DCARS gene pair: SKP1 and SKP2
# calculates p-value based on permutation
DCARS(SKCM_rank,"SKP1","SKP2",plot=TRUE)
#> weight matrix not specified, generating weight matrix now
#> span not specified, defaulting to 0.5
#> performing permutation test
#> [1] 0
# extract only the test statistic
DCARS(SKCM_rank,"SKP1","SKP2", extractTestStatisticOnly = TRUE)
#> weight matrix not specified, generating weight matrix now
#> span not specified, defaulting to 0.5
#> [1] 0.227379

# not significantly DCARS gene pair: EIF3C and EIF5B
# calculates p-value based on permutation
DCARS(SKCM_rank,"EIF3C","EIF5B",plot=TRUE)
#> weight matrix not specified, generating weight matrix now
#> span not specified, defaulting to 0.5
#> performing permutation test
#> [1] 0.97
# extract only the test statistic
DCARS(SKCM_rank,"EIF3C","EIF5B", extractTestStatisticOnly = TRUE)
#> weight matrix not specified, generating weight matrix now
#> span not specified, defaulting to 0.5
#> [1] 0.01790925

# build weight matrix
W = weightMatrix(ncol(SKCM_rank), type = "triangular", span = 0.5, plot = TRUE)
#> Loading required package: gplots
#> 
#> Attaching package: 'gplots'
#> The following object is masked from 'package:stats':
#> 
#>     lowess

# extract DCARS test statistics
SKCM_stats = DCARSacrossNetwork(SKCM_rank,edgelist = STRING,
                                W = W, extractTestStatisticOnly = TRUE,
                                verbose = FALSE)
sort(SKCM_stats,decreasing=TRUE)[1:10]
#>    GMPS_IMPDH1      SKP1_SKP2     COPB2_COPE ANAPC10_ANAPC5     TBL3_UTP18 
#>      0.2518867      0.2273790      0.2075771      0.1800155      0.1672255 
#>   PSMD12_PSMD3  POLR3B_POLR3D    RAD18_UBE2B    SNRPB_SNRPG   RPL26L1_RPL8 
#>      0.1659466      0.1629938      0.1621321      0.1555048      0.1531930