Interpret results about sierra

[rj-patrick]

Hi @Vivianstats,

1. Yes, the coordinates correspond to start-end coordinates of the peak
2. The two columns refer to the % of cells expressing the peak in your target and comparison cell populations, respectively (documentation for DUTest has been updated now to explain this).
3. The Log2 fold change represents the relative difference in peak expression between the two groups of cells. It is derived from the DEXSeq estimateExonFoldChanges function which uses a GLM to estimate relative fold change change differences between conditions.
4. If you are interested in looking at 3'UTR shortening, you can use the DetectUTRLengthShift function for that - see the details in the Vignette or on the wiki.

Cheers
Ralph

[Vivianstats]

Hi Ralph,

Thank you very much for the clarification! I have an additional question:
Are the overall gene counts summarized somewhere in Sierra's results? For each gene, do its peak counts sum up to the total read count?

[rj-patrick]

We don't currently include the overall gene counts. If you're comparing to something like CellRanger or STARsolo the summed peak counts across all genomic feature types will be different - bear in mind that Sierra calls peaks across the gene body, which includes introns. The 3'UTR -specific peak counts will have the highest correlation with gene counts but won't perfectly match.

[Vivianstats]

Thanks for the information!