sdRain {twoddpcr} | R Documentation |
Although we can use various algorithms to classify all droplets in a ddPCR experiment, there will be some variation between the classifications. We can perhaps have a relatively high confidence that droplets near the centres of clusters do indeed belong to that cluster, whereas we probably have a lower confidence in the classification of those further away, say, near the 'boundary' of two clusters. We may view these droplets (or a subset of them) as having an ambiguous class. This function allows us to only consider droplets classified within a certain distance of the means of each cluster and label the rest as "Rain".
sdRain(droplets, cMethod, errorLevel = 5, ...) ## S4 method for signature 'data.frame' sdRain(droplets, cMethod, errorLevel = 5, fullTable = TRUE) ## S4 method for signature 'ddpcrWell' sdRain(droplets, cMethod, errorLevel = 5) ## S4 method for signature 'ddpcrPlate' sdRain(droplets, cMethod, errorLevel = 5)
droplets |
A |
cMethod |
The name or column number of the classification for which we want to add rain to. |
errorLevel |
How many multiples of standard deviation from the mean of each cluster to retain. Can be a list where each item corresponds to a class name and the multiple for that class. Can also be a numeric vector of length 1, which is equivalent to a list with all the same entries. Defaults to 5. |
... |
Other options depending on the type of |
fullTable |
If |
If droplets
is a data frame, return a data frame or factor
(depending on fullTable
) where droplets with ambiguous
classifications are labelled as "Rain".
If droplets
is a ddpcrWell
object, return
a ddpcrWell
object with a rainy classification.
If droplets
is a ddpcrPlate
object, return
a ddpcrPlate
object with a rainy classifications.
Anthony Chiu, anthony.chiu@cruk.manchester.ac.uk
This approach was described in Jones, M., Williams, J., Gaertner, K., Phillips, R., Hurst, J., & Frater, J. (2014). Low copy target detection by Droplet Digital PCR through application of a novel open access bioinformatic pipeline, "definetherain." Journal of Virological Methods, 202(100), 46–53. http://doi.org/10.1016/j.jviromet.2014.02.020
## Compare the types of droplets in a single well for the "Cluster" class ## and then with rain. aWell <- ddpcrWell(well=KRASdata[["E03"]]) aWell <- sdRain(aWell, cMethod="Cluster") cl <- wellClassification(aWell) table(cl$Cluster) table(cl$ClusterSdRain) ## Compare the types of droplets in multiple wells for the "Cluster" class ## and then with rain. krasPlate <- ddpcrPlate(wells=KRASdata[c("E03", "H03", "C04", "F04")]) krasPlate <- sdRain(krasPlate, cMethod="Cluster") plateSummary(krasPlate, cMethod="Cluster")[, c(1:5)] plateSummary(krasPlate, cMethod="ClusterSdRain")[, c(1:5)] ## The 'errorLevel' parameter can changed. krasPlate <- sdRain(krasPlate, cMethod="Cluster", errorLevel=4) plateSummary(krasPlate, cMethod="ClusterSdRain")[, c(1:5)] ## The 'errorLevel' parameter can also be changed for each cluster. krasPlate <- sdRain(krasPlate, cMethod="Cluster", errorLevel=list(NN=5, NP=5, PN=4, PP=3)) plateSummary(krasPlate, cMethod="ClusterSdRain")[, c(1:5)]