How to generate GFF?

[ykdang]

I am working on a fungus call Neurosprora. Previous lab member built a gbrowser (1.70) to visualize the deep-seq data with GFF2 file format. Now I obtain some new deep-seq data and want to load it onto that browser. I heard that he used bowtie to do alignment and convert into gff2, but I cannot find the script to convert. Does anyone have similar script which could share? Or just indicate any tools can do this trick? I am not a perl person and don't think I could learn it in a short while.

[gringer]

If you have a BAM file (which bowtie can produce), this tutorial suggests that GBrowse can handle BAM directly:

http://gmod.org/wiki/GBrowse_NGS_Tutorial#Tell_GBrowse_About_the_SAM_Files

You could also have a look at the admin tutorial:

http://gmod.org/gbrowse2/tutorial/tutorial.html#graph

[gringer]

FWIW, I've just today ended up writing a python script to convert from SAM/BAM files to GFF3 files using pysam. The code may be useful for you if you can't find anything else suitable for your conversion.

For a bit of context, I broke up genomic contigs into 100bp fragments, naming the sequences <contig>#<start>-<end>, then used bowtie2 to map them to the genome -- that's why I've got the 'read.qname.find' bits in the code. My contigs started with 'v', which bowtie replaced with 'N' for some odd reason, so I had to do a bit of extra fiddling to add the 'v' back in.

Here's the relevant part of my code which does the SAM->GFF conversion:

samFile = pysam.Samfile(samFileName, "r")
totalCount = 0
sys.stderr.write("Getting reads from SAM file...")
sys.stdout.write("##gff-version 3\n")
gffWriter = csv.writer(sys.stdout, delimiter = '\t')
for read in samFile:
    totalCount += 1
    if(read.tid > 0):
        qContig = 'v' + read.qname[1:read.qname.find("#")]
        qContigStart = read.qname[read.qname.find("#")+1:read.qname.find("-")]
        qContigEnd = read.qname[read.qname.find("-")+1:]
        tContig = samFile.getrname(read.tid)
        strand = "-" if read.is_reverse else "+"
        score = "."
        for tag in read.tags:
            if(tag[0] == "XS"):
                score = str(tag[1]+1000)
        gffWriter.writerow((qContig,
                            "sam2gff3-"+os.path.basename(samFileName),
                            "nucleotide_match", qContigStart, qContigEnd,
                            score, strand, ".",
                            "Name=SAM_%s,ID=%s-%s;Target=%s %d %d" %
                            (tContig, qContig, tContig, tContig,
                             read.pos, read.aend)))
    if(totalCount % 100000 == 0):
        sys.stderr.write(".")

[ykdang]

Thanks a lot, that's could be very useful