Is there a good way to get the proportions of isoforms *within* a condition using output from kallisto

Question

[brett.vanderwerff]

I am using kallisto to analyze some RNA-seq samples.

So I know that you can get a good idea of the differential expression of isoforms across conditions downstream of kallisto using tools like sleuth and IsoformSwitchAnalyzeR. What I am wondering is if there is method or library for determining the proportion of isoforms within a condition, not across conditions.

Lets say my favorite gene has 3 isoforms (A, B, and C)

Lets say I have output from kallisto from 3 replicates of samples not treated with anything. I am really interested in knowing the proportion of A, B, and C present in this baseline condition.

Is there a good way to figure out, say for example that isoform A is the most abundant isoform in my 3 untreated samples?

Thank you.

[Kristoffer Vitting-Seerup]

Could you not just for each gene calculate the mean isoform expression over your 3 replicates and then look for which is the largest?

[brett.vanderwerff]

I wasn't sure because the isoforms might extract different or have different lengths which would affect the numbers of reads they get vs some other isoform from the get go. I know cross gene comparisons are kind of tricky so I thought I would ask.

[Kristoffer Vitting-Seerup]

Sure - you would have to use length normalized expression values such as the TPM estimated by Kallisto