crosshyb {crosshybDetector} | R Documentation |
This function identifies probes that are potentially affected by cross-hybridization.
The probability of a probe of being a corruptor is computed by Monte Carlo simulations.
The algorithm is briefly described into the crosshyb_func
help page.
crosshyb(raw, probeSeq, plate = 1, probeNameID = "ProbeName", numPermut = 10000, probes = c("probes", "spike"), satValue = 65535, maxProbes = 50, delta = 10)
raw |
an object of class marrayRaw |
probeSeq |
a vector with the probe sequences |
plate |
integer, the array to analyze |
probeNameID |
the slot name of raw containing the probe names.
Probes names are extracted with maInfo(maGnames(raw))[[probeNameID]] |
numPermut |
integer, the number of permutations for Monte Carlo simulations |
probes |
the type of probes that are analyzed for potential corruption.
Probe types are defined in maControls(raw) |
satValue |
integer, the value corresponding to saturation (default 65535) |
maxProbes |
integer, the maximun number of probes to analyze for potential corruption |
delta |
integer, threshold for difference in melting temperature. Pairs of probes whose delta Tm
is lower than delta are considered as similar. Delta Tm i,k is computed as Tm i,i - Tm i,k,
where Tm i,i is the Tm of the perfect pairing probe i vs probe i and Tm i,k is the Tm of the
imperfect pairing probe i vs probe k
|
A list containing:
dataR |
Dataframe with three columns containing the results of the analysis for the putative corruptor probes applied to the red channel. Contains: i) the probe number, ii) the probe name and iii) the pvalue based on Monte Carlo simulations of being a corruptor (pvalue is corrected for multiple testing) |
dataG |
Same as dataR , but applied to the green channel |
childrenR |
A list containing the corrupted probes for each putative corruptor from the red channel |
childrenG |
As childrenR , but applied to the green channel |
Paolo Uva
crosshybMCplot
, extractBadProbes
, crosshyb_func
## Not run: data(raw) data(probeSeq) # This function will take several minutes to finish crosshyb.out <- crosshyb(raw, probeSeq, plate=1, numPermut=10000, probeNameID="Name", probes=c("probes", "spike"), satValue = 65535, maxProbes=100) ## End(Not run)