This is an old revision of the document!
The NOISeq tutorial can be downloaded from Bioconductor website. Click here.
In NOISeq package, you can find not only the method to compute differential expression between two experimental conditions from RNA-seq data, but also other functions in order to learn more about saturation, contamination or other biases in your data as well as to explore the differential expression results.
Please, follow these instructions to run NOISeqBIO:
1) Install NOISeq Bioconductor package (see Downloads) and load it: library(NOISeq)
.
2) Download the zip file containing the R scripts for NOISeqBIO (from Downloads) and unzip it.
3) To load the NOISeqBIO functions in R from your working directory:
# Remember you have to change this path if you are not in the same directory: > mypath = "NOISeqBIO_v00" > source(file.path(mypath, "auxiliar.R")) > source(file.path(mypath, "MDbio.R")) > source(file.path(mypath, "allMDbio.R")) > source(file.path(mypath, "noiseqbio.R")) > source(file.path(mypath, "classes.R")) > source(file.path(mypath, "normalization.R")) > source(file.path(mypath, "fewreplicates.R"))
4) Read your data as in NOISeq (see the following example with Marioni's data, where liver and kidney tissues are compared) to convert them to a NOISeq object:
> data(Marioni) > mydata = readData(data = mycounts, factors = myfactors)
5) To apply NOISeqBIO with TMM normalization (you can choose the same normalization procedures as in NOISeq):
> mynoiseq.test = noiseqbio(mydata, norm = "tmm", factor = "Tissue", r = 10)
The parameter r is the number of permutations of sample labels to generate null distribution.
6) Finally, select the differentially expressed genes:
# Probability cutoff. It is equivalent to a FRD (adjusted p-value) of 0.05 > myq = 0.95 > mydeg = mynoiseq.test@results[[1]][which(mynoiseq.test@results[[1]][,"prob"] > myq),] > nrow(mydeg) [1] 2837 > head(mydeg) Kidney Liver theta prob ENSG00000187642 12.8606633 4.332902 0.8107475 0.9601231 ENSG00000188290 17.7544458 4.961376 1.3160369 0.9999999 ENSG00000187608 19.2702201 34.329320 -0.7153955 0.9543329 ENSG00000188157 691.1874857 141.575959 5.9237953 1.0000000 ENSG00000186891 0.8230924 2.736435 -0.7348520 0.9568449 ENSG00000078808 372.5016774 483.708484 -1.0066939 0.9787270