Locating nearest genes in RRBS analysis
2
0
Entering edit mode
Mageiros • 0
@f35f12a1
Last seen 9 weeks ago
Greece

I am trying to analyze a data set of BS-seq (RRBS) data following the example at the section 4.7 of the edgeR user manual.

In the tutorial the function nearestTSS is used to annotate the nearest gene of each site.

Is there a way to retrieve information though for the two upstream and two downstream genes instead of the one that is closer?

I found the get2NearestFeature function (https://rdrr.io/bioc/hiAnnotator/man/get2NearestFeature.html) but I am not sure how to convert my DGElist object to a GRanges objects.

Note that I am using hg19 as a human reference strain.

Best Leonardos

0
Entering edit mode
@gordon-smyth
Last seen 50 minutes ago
WEHI, Melbourne, Australia

You can form a GRanges object from the vector of chromosome names and the vector of locations, same as you input to nearestTSS. The help page you link to gives an example of how to do that. I'm not familiar with get2NearestFeature though so can't help you with it.

Regarding nearestTSS, please be aware that it only works with hg38, not with hg19, so you should not use it at all in your analysis.

Is there a reason for using hg19, since hg38 was released nearly a decade ago?

0
Entering edit mode
Mageiros • 0
@f35f12a1
Last seen 9 weeks ago
Greece

Thx a lot for your reply. This project was handed to me with the bismark results already using hg19. I did everything from the beginning using hg38. I will try now to use get2NearestFeature. I will let you know how it works if a manage to do it successfully.

Best Leonardos