RNA-seq reads mapping to coding regions

[m!x]

Hi,

I am trying to analyze my RNA-seq data and would like to know how many of the reads mapped to just the coding regions in the genome.

I have previously mapped the reads to UCSC mouse reference genome with Bowtie.
I mainly followed the protocol in Barbara Wold's paper (ERANGE) to do the subsequent analysis, i.e. creating RDS file from the BOWTIE-mapped reads using UCSC knownGene table.
Everything went quite well and I got the final RPKM results.

I am currently trying to use the BOWTIE-mapped reads and convert it to RDS format, but using CCDSgene table (which supposedly only contain the coding regions) instead of knownGene table. However, I always got this error message:

Traceback (most recent call last):
File "/net/shendure/vol1/home/lilyanam/commoncode/makerdsfrombowtie.py", line 114, in <module>
rds = readDataset(outdbname, init, dataType, verbose=True)
File "/net/shendure/vol1/home/lilyanam/commoncode/commoncode.py", line 652, in __init__
self.initializeTables(self.dbcon)
File "/net/shendure/vol1/home/lilyanam/commoncode/commoncode.py", line 761, in initializeTables
acon.execute('create table metadata (name varchar, value varchar)')
sqlite3.OperationalError: table metadata already exists


Have anyone tried this strategy before to count the reads that mapped to coding regions only?
or, is there any better way to do this?

Thank you very much for the suggestion!

-Lilyana