Cage-Seq Data Analysish

Question

[dli]

Hi,

Is anyone have experience of CAGE-seq data analysis? Currently I did basic mapping to genome (Am I right that mapping to genome?). I have following specified questions actually:

1, do I need remove redundant mapped tags, i.e., tags mapped to same locations?

2, how did I do the normalization? I have search some related paper, the formula were too complex to me through

3, how did I find top-enriched regions? I suspect most if the regions should around promoter region and could use the existed annotation to found them, but this way I might lose some new promoter regions?

Any response would be greatly appreciated. Thanks.

Answer 1

[Mikael Huss]

On (1), I don't think you should remove redundant mapped tags, because you will expect a high density of tags mapping to the same location due to the way the technique CAGE works with lots of enrichment at the 5' end. On the rest I think you just need to read the CAGE papers to find out what they did.

EDIT: This R package maybe? CAGEr

[dli]

Nice suggestions! I will try :)

Answer 2

[ancient_learner]

As CAGE also produces results similar to RNA-SEQ experiments i think u can use packages like EdgeR or others to normalize the data. first you have to use your mapped reads to count their occurrences.

[dli]

Thanks for reply. I think RNA-seq data was usually mapped to transcriptome instead of genome? Should I mapped the tags to cDNA then?