I am using Tophat on some Solid fungal transcriptomes and want to use the junctions found by Tophat as intron hints in the Augustus gene finder. In the junctions.bed file produced, the coordinates for the junctions (columns 2 and 3) seem to not only include the intron but also parts of the flanking exons, and to get the intron coordinates I have to add or subtract the BlockSizes values given in column 11 in the bed file.
Example.
junctions.bed:
ctg220009118880 675 839 JUNC00004318 14 + 675 839 255,0,0 2 37,38 0,126
If correctly converted to GFF3 format for Augustus:
ctg220009118880 tophat intron 713 801 0 . . mult=14;src=E
The third and fourth column of the GFF format are the coordinates converted from column 2 and 3 +/- column 11 in the BED file taking into account that bed files start with coordinate 0, GFF with 1 (start:675+37+1=713, stop:839-38).
I can write a script that recalculates the coordinates and converts the bed file into a gff3 format that Augustus wants, but I cannot help but feel that I am missing something. Are there any existing solutions out there? Am I simply misunderstanding something?
All suggestions are very much appreciated.
Example.
junctions.bed:
ctg220009118880 675 839 JUNC00004318 14 + 675 839 255,0,0 2 37,38 0,126
If correctly converted to GFF3 format for Augustus:
ctg220009118880 tophat intron 713 801 0 . . mult=14;src=E
The third and fourth column of the GFF format are the coordinates converted from column 2 and 3 +/- column 11 in the BED file taking into account that bed files start with coordinate 0, GFF with 1 (start:675+37+1=713, stop:839-38).
I can write a script that recalculates the coordinates and converts the bed file into a gff3 format that Augustus wants, but I cannot help but feel that I am missing something. Are there any existing solutions out there? Am I simply misunderstanding something?
All suggestions are very much appreciated.
Comment