RNA-seq read coverage plot for genes?

[asiangg]

Hi all:

I'm doing some data crunching on our lab's RNA-seq data and have encountered a difficult task. One thing I would like to do is to plot the averaged read coverage for a group of genes. It sounds similar to a promoter plot but is complicated by the nature of discontinuous exons and various transcript lengths. Not sure how this can be done. Can anyone refer to a software tool or method? Does anyone understand what I'm talking?

[steven]

I guess you want to compute the read coverage along transcripts (after splicing, so not genes), like Levin et al. did in this nice paper to assess the evenness of the coverage. Is that correct?
I do not know any publicly available tool doing that. let me know if you find one!

[natstreet]

I'm not sure if I've understood exactly what you want but have a look at this GBrowse example to see if it's anything like what you want. It was made using the output from tophat. If it's of any use I can detail the steps etc to do the same.

[asiangg]

Steven,
Yes. Fig. 4b is the plot that I'm trying to replicate for our data. However, it seems most authors do not detail the procedure to do that.

natstreet,
Nice example but getting coverage plot for individual genes is not my major concern. I'm puzzled by how to average many plots and get a global taste of how it looks like.

Thanks!

[jdanderson]

@Natstreet- Hello Natstreet,

I would be very interested in reading your protocol for displaying Gbrowse data, as I am not familiar with Gbrowse. If it is at all possible, could you also post any help guidelines for entering CuffDiff data files into Gbrowse.

Thank you for your generous offer, I look forward to hearing from you.

Regards,
Johnathon

[perrinwang]

I am also interested in knowing a way to show such average gene coverage plot... is there any software/script out there that can do this based on alignment files (SAM, BAM) and an gene annotation file (GFF, GTF)?

[lexa]

has anyone found a solution to this problem?

[zorph]

i am going to eventually do something similar for the data in my lab as well. I cam across this webiste from illumina's TRU-SEQ.

http://www.illumina.com/truseq/tru_resources/scripts.ilmn

it contains scripts that will help in calculating a normalized coverage histogram and a coverage gap histogram. I'm sure you can modify this to look at groups of genes seeing as they chose to do their example with chr21.

Good luck.

[lexa]

Thanks a lot! it seems to be what I am looking for.