UMI counts and read counts in GeneGlobe output

Question

[Abhishek]

Hi all,

I have obtained the output file from QiaSeq GeneGlobe containing UMI counts. The description of quantification procedure is given here : https://www.qiagen.com/us/resources/resourcedetail?id=bea2dcfa-0a5c-47c5-afd8-8b0fe90a471a&lang=en

Subset of the data in the output excel file from 'miR_piRNA' sheet is given below

I'm not exactly sure why there are both UMI counts and read counts in the output file.

My assumption is that the READs columns are the initial counts obtained and then after collapsing the duplicate UMIs, the counts are given in UMIs columns. Is this correct ?

If so, then for further analysis - i.e. DE analysis etc, only the UMI columns are to be used ? Is there any use of the READs columns ?

Answer 1

[i.sudbery]

Your understanding is correct. The reads are the counts of all reads mapping to a given gene. UMIs is the number of distinct UMIs, and therefore the number of original RNA fragments sampled before PCR was applied. Downstream analysis should use UMIs.

[Abhishek]

Thank you for the confirmation.

Would there be any reason the READs columns are provided too ? I'm thinking maybe for normalizing the UMIs columns for further downstream analysis, the READs columns should be used somehow ?