Isoform/switch/gene ratio

[zoe]

Hi Kristoffer,

I have a question about the definitions of nrIsoforms, nrSwitches, and nrGenes. 

The extractSwitchSummary function showed that I have 1835 isoforms, 1568 switches, and 1070 genes.

Comparison nrIsoforms nrSwitches nrGenes

1 X13 vs X24       1835       1568    1070

As you can see, I have fewer isoforms than 2x genes. If nrGenes is the number of differentially expressed genes between my two samples, why aren't there at least two isoforms for each gene? Similarly, how are there more switches than there are genes - I thought with two conditions each gene could only 'switch' once. 

I'm sure I'm just misinterpreting what defines a switch, isoform, and gene. Any clarification on exactly how the program defines each of these terms would be helpful.

Thanks in advance,

Zoe

[Kristoffer Vitting-Seerup]

Hi Zoe

Thanks for reaching out.

The reason the numbers does not match are that each category is defined differently:

- nrIsoforms: is defined as the number of isoforms which are differentially used.

- nrSwitches: is defined as the number of switches identified. A switch consist of one isoform which is differentially used an another isoform with an opposite change in usage. Using default setting this opposite change does not itself have to be significant but just be of a particular effect size. You can change this in most IsoformSwitchAnalyzeR functions using "onlySigIsoforms" argument.

- nr genes is simply the number of genes which have at least one differentially used isoform.

And you can easily have multiple switches per gene.

Let me know if this helped.

Cheers

Kristoffer