How to compulate the FPKM of differentially expressed transcripts (DETs)

Question

[gengyang12138]

I'm study the Alternative splicing events of RNA-seq. I saw a artical, which said:

It was worth mentioning that some splicing transcripts of a certain gene were differentially expressed in different libraries. We called these genes differentially expressed transcripts (DETs). And then, author compulated the expression (FPKM) of all splice transcripts of (DETs)

Firgure.1 part of the Splicing transcripts and gene structures, becaues of the size of picture, I just pasted a part of the picture.

Figure.2 The FPKM of splice transcripts.

I don't know how to compulate the FPKM of splice transcripts, or which software can calculate the values. If possiable, I also want to know how to draw the pictures like Figure. 1

[WouterDeCoster]

Not what you are asking for but FPKM is generally not a good method for normalization. Why do you want it?

[lieven.sterck]

and no idea where it came from but I see this appearing more and more in papers (where they indeed use this for differential expression analysis ????)

Answer 1

[Kristoffer Vitting-Seerup]

It might help if you link the acutal paper - but here is what I think is going on:

There are two parallel analysis going on:

1. Quantification of transcript data (resulting in FPKM values). Loads of tools for that - see my discussion here.
2. A toolwhich detect differential splicing. Also loads of tools for that.

The neat part is that if there is alternative splicing that will result in a different transcript - which means that you can identify the whole transcript that is changing.

Tools for doing the quantification can be found via the link above. A nice tool for analysing the result (and consequence) of alternative splicing is my R package IsoformSwitchAnalyzeR which can both identifying and visualize such isoform switches (both individual and genome wide). Examples of visualization can be seen here.

[gengyang12138]

Thank you very much. The artical is here enter link description here. And I think your R package can reslove my problems very well.

[lieven.sterck]

The neat part is that if there is alternative splicing that will result in a different transcript - which means that you can identify the whole transcript that is changing.

Keep in mind though that those differences can be very small (only couple of bases) and thus the alternative transcripts can be nearly identical, which makes that identifying them as DET can become very hard to impossible.

+1 for the R-package

Answer 2

[hafiz.talhamalik]

You calculate RPKM/FPKM using simple three steps. 1-Count up the total reads in a sample and divide that number by 1,000,000 – this is our “per million” scaling factor. 2-Divide the read counts by the “per million” scaling factor. This normalizes for sequencing depth, giving you reads/fragments per million (RPM/FPM) 3-Divide the RPM values by the length of the gene, in kilobases. This gives you RPKM/FPKM.

[hafiz.talhamalik]

to differentiate between RPKM/FPKM see this video. its quite helpful

[gengyang12138]

Yes, I knew how to calculate the FPKM of genes, but, in this artical, the author calculated the expression of differential transcripts.

In other words, I want to know if there some software can analysis the alternatice splicing and calculate the expression of transcripts of one genes.