Question: RNASeq biplot variables and samples
0
5 weeks ago by
ecg1g150
ecg1g150 wrote:

I would like to generate a biplot

I have been following the Love et al (2019) version of RNASeq analysis using DESeq, and have been able to plot PCAs, however, they haven't designed a function of generating billets yet

http://bioconductor.org/packages/devel/bioc/vignettes/DESeq2/inst/doc/DESeq2.html

DO you know how to transfer the PCA data from here to a biplot package?

I have checked frizzpcavar from the "facto extra" package but I am not able to input the variables, and is taking the gene IDs as variables (therefore a very messy graph).

DO you have other recommendation which might be easier to use?

My data looks like this (values are random)

Gene_name Sample1 Sample2 Sample 3 Sample4
Gene1 452435 23434 2342345 234235
Gene2 2435253 235245 56767 64365
gene3 4352345 56793 8437 46567
Gene4 9834757 98745 63404 92875


Environmental data looks like this

Samples Variable1 Variable2 Variable3 Variable4
Sample1 0.2 23 0.5 9 1
Sample2 0.2 23 0.5 9 2
Sample3 0.4 24 0.8 10 2
Sample4 0.3 24 0.7 10 1


Thanks

rnaseq deseq2 • 82 views
modified 5 weeks ago by Michael Love25k • written 5 weeks ago by ecg1g150
Answer: RNASeq biplot variables and samples
0
5 weeks ago by
Michael Love25k
United States
Michael Love25k wrote:

I'd recommend you take the transformed data itself and use another package to perform the dimension reduction, if you want a particular plot.

E.g., use:

mat <- t(assay(vsd))


As your matrix for dimension reduction. I've transposed it so that samples are now rows (typical for base R functions).