This function allows to normalize repertoires using the Shannon entropy as a threshold, thus eliminating rare sequences.
It is described in (Chaara et al., 2018) and is used to eliminate “uninformative” sequences resulting from experimental noise.
It is particularly efficient when applied on small samples as it corrects altered count distributions caused by a high-sequencing depth.
Sequences that are eliminated with this method are stored in the oData
slot.
Arguments
- x
an object of class
RepSeqExperiment
Value
a new RepSeqExperiment object with the
normalized data.
Examples
data(RepSeqData)
RepSeqData_sh<- ShannonNorm(x = RepSeqData)
#> Creating a RepSeqExperiment object...
#> Done.