PCA plot- R/Bioconductor-Deseq2

Question

[Lindsay44]

I am trying to make a PCA plot using R/Bioconductor- Deseq2. I have RNA-seq data saved to my computer. How to I upload this data to the program and what commands do I use to create the plot?

I am using this visualization method to see if my two experimental groups are compatible before I enter them into Cuffdiff. Which files do I use as input (raw fastq files or the SAM files from the aligner)?

I am also interested in making a differential analysis plot using the same programs when I am done running the data through Cuffdiff. What commands are used for this?

Thanks for the help.

[GouthamAtla]

http://www.bioconductor.org/help/workflows/rnaseqGene/

[WouterDeCoster]

You'll need read counts for creating a PCA plot, generated by HTSeq-count of featureCounts for example. For the rest of your questions, do you know a bit of R?

Answer 1

[karl.stamm]

You're asking for an awful lot of automation. Maybe you would have good results trying my software made for this purpose.

Rgsepd is a tool that takes your read-counts matrix and sample names, then performs DESeq2 for you, and makes the PCA plots too. It requires the read-counts matrix to have human RefSeq transcript identifiers as the rows. Let me know if that's a problem.

http://bioconductor.org/packages/3.4/bioc/html/rgsepd.html

The tool continues to do a few more steps, but you could ignore them if you only want the PCA images. The full name of rgsepd is "Gene Set Enrichment and Projection Displays" because it will also perform GOSeq to tell you the interesting functional sets present in your data; then calculate if and how much each functional set differentiates your samples.

[Lindsay44]

That sounds great; however, my data is from mice. Will it still work? Thanks.

[karl.stamm]

I should give that a try. I suspect it will get past DESeq2 then fail while trying to convert the gene identifiers for the GOSeq portion. You might still get the PCA output. And if not, I should update the code to do so.
Which kind of identifiers do you use?

[Lindsay44]

I am not completely sure- but I know the data was sequenced using an Illumina sequencer. I am trying to find ways to visualize my data so in addition to the PCA plot I will also be looking to make heat maps, etc. what kind of files should be used for the PCA plot?