deseq2 user defined size factors

Question

[Damian Kao]

Is it possible to skip the estimateSizeFactor step in Deseq2 and define your own size factors? If I just replace the size factor vector with my own values, will the down-stream DE steps use these values? I am pretty new with R so it would be very helpful if I can see some example code on how to do this.

Answer 1

[Carlo Yague]

Hi,

Short answer : yes and yes !

As for the example code, see my answer here.

or more to the point :

sizeFactors(cds) = c(my_Values)

[rodd]

How is this then passed onto another dds object? (I am trying to analyse only a subset of genes from all genes expressed in a dataset because I am focusing on genes which are modulated by eQTLs - but ideally I would like to normalise raw counts to the size factors calculated using the whole dataset to improve the normalisation protocol, right?)

[rodd]

Ooops. I guess it's better I don't do that, actually. Looks like it's better to just use the whole dataset, and perform multiple testing correction on the list of genes I am interested, as it seems tricky to stipulate size factors from one dds object to another. Source: https://support.bioconductor.org/p/79470/#79477

Answer 2

[ivivek_ngs]

Yes you can, that is what I can read from the below link

Take a look here and here

There was a post from where Damian you already replied few years back where I can see Jeremy replied that you can set the size factor manually as well with sizeFactors

dds <- estimatesizefactors(dds)<="" code=""> See DESeq for a description of the use of size factors in the GLM. One should call this function after DESeqDataSet unless size factors are manually specified with sizeFactors. Alternatively, gene-specific normalization factors for each sample can be provided using normalizationFactors which will always preempt sizeFactors in calculations. Internally, the function calls estimateSizeFactorsForMatrix, which provides more details on the calculation.