Creating a gtf file for the database I mapped against

[lanner]

I used BWA-SOLiD from Galaxy, but had somewhat of a mapping failure with very little hits. It was kindly suggested to me by Jennifer of Galaxy:

"I would consider the mapping a failure. However, you can create a gtf file for the database you mapped against and generate counts for the hits you have. This older post on the UCSC forum has instruction (the file cannot be output from the UCSC Table browser, so cannot be done through the Galaxy "Get Data -> UCSC Main" tool).
http://redmine.soe.ucsc.edu/forum/index.php?t=msg&goto=14375&S=7744d556a73fd27a61976e12db2bc65c"

I tried to follow the advice outlined on that website. Specifically, I

1) Installed bedToGenePred, bedFile, and genePredFile

2) Ran the command:

cat mrna.fasta | cut -f2- | pslToBed stdin stdout | bedToGenePred stdin stdout | genePredToGtf file stdin mrna.gtf

But I got an error "Bad line 1 of stdin wordCount is 2 instead of 21 or 23".

I wrote to the group about it, but was wondering if anyone on SeqAnswers might know either what is wrong in this specific case, or if there is another easy-to-use tool out there to accomplish what Jennifer suggested (to create a gtf file for the database I mapped against).

Many thanks in advance!

This is the head of my mrna.fasta file:

>AF001540 1
ggcacgaggcaggtctgtctgttctgttggcaagtaaatgcagtactgtt
ctgatcccgctgctattagaatgcattgtgaaacgactggagtatgatta
aaagttgtgttccccaatgcttggagtagtgattgttgaaggaaaaaatc
cagctgagtgataaggctgagtgttgaggaaatttctgcagttttaagca
gtcgtatttgtgattgaagctgagtacatttgctggtgtatttttaggta
aaatgcttttttgttcatttctgggtggtgggaggggactgaagccttta
gtcttttccagatgcaaccttaaaatcagtgacaagaaacattccaaaca
agcaacagtcttcaagaaattaaactggcaagtggaaatgtttaaacagt
tcagtgatctttagtgcattgtttatgtgtgggtttctctctcccctccc

[mikep]

Which database did you map against? if it's a common one then the gtf is probably readily available (and I'd use gencode over UCSC if it's human or mouse, but that's just me).

At any rate, the instructions you are using are for reformatting annotation in psl format to gtf. You seem to be piping actual sequence into the pipeline, which is why it failed. You want to create a gtf file from the annotation for your database, not the sequence.