Hi guys, 

I'm new to bioinformatics so this may be a naive question. My workflow:

1.  tximport function to create the txi data frame from .h5 kallisto files
2.  DESeqDataSetFromTximport function to generate DESeq data set

Question:

Do I need to generate gene-level unnormalized counts during the import*? or just to use transcript-level counts**?

---

I found the article "Importing transcript abundance datasets with tximport" very helpful but still got confused with the following paragraph:

Note: there are two suggested ways of importing estimates for use with differential gene expression (DGE) methods. The first method, which we show below for edgeR and for DESeq2, is to use the gene-level estimated counts from the quantification tools, and additionally to use the transcript-level abundance estimates to calculate a gene-level offset that corrects for changes to the average transcript length across samples. The code examples below accomplish these steps for you, keeping track of appropriate matrices and calculating these offsets. For edgeR you need to assign a matrix to y$offset, but the function DESeqDataSetFromTximport takes care of creation of the offset for you. Let’s call this method “original counts and offset”. 

Thanks in advance,
Yunlu

*

txi.kallisto.gl <- tximport(files, type = "kallisto", tx2gene = tx2gene)

**

txi.kallisto <- tximport(files, type = "kallisto", txOut = TRUE)​

Here’s my comment to this Q on another thread:

C: Incredibly high/low foldChange

To repeat from that thread, it works and will find DE at the transcript level, but you should consider to apply a stricter padj threshold. And we are meanwhile working on improvements.