computeSD.func
From aCGH v1.50.0
by Peter Dimitrov
Function to estimate experimental variability of a sample
This functions estimate experimental variability of a given sample. This value can be used to rank samples in terms of the quality as well as to derive thresholds for declaring gained and lost clones.
- Keywords
- models
Usage
computeSD.Samples(aCGH.obj, maxChrom = 22, maxmadUse = .3, maxmedUse = .5, maxState = 3, maxStateChange = 100,
minClone = 20)
computeSD.func(statesres, maxmadUse = 0.2, maxmedUse = 0.2, maxState = 3, maxStateChange = 100, minClone = 20, maxChrom = 22)Arguments
- aCGH.obj
-
Object of class
aCGH. - statesres
- The states.hmm object, generally is the output of
mergeFunc. - maxmadUse
- Maximum median absolute deviation allowed to controbute to the overall variability calculation.
- maxmedUse
- Maximum median value for a state allowed to contribute to the calculation.
- maxState
- Maximum number of the states on a given chromosome for the states from that chromosome to be allowed to enter noise variability calculation.
- maxStateChange
- Maximum number of changes from state to state on a given chromosome for that chromosome to enter noise variability calculation.
- minClone
- Minimum number of clones in a state for clones in that sate to enter variability calculation.
- maxChrom
- Maxiumum chromosomal index (generally only autosomes are used for this calculation.
Details
Median absolute deviation is estimated in all the states passing the criteria defined by the parameters of the function. Then median of all MADs on individual chromosomes as well as across all chromosomes is taken to estimate chromosomal experimental variability and sample experimental variability.
Value
- madChrom
- Returns a matrix containing estimated variability for each chromosome for each sample.
- madGenome
- Returns a vector with estimate of experimental varibility for each sample.
References
Application of Hidden Markov Models to the analysis of the array CGH data, Fridlyand et.al., JMVA, 2004
See Also
Community examples
Looks like there are no examples yet.