Searching for introns within 5'UTRs

Question

[ryan.borotra]

Hi everybody!

I am looking for the best way to find the sequences of introns that are located within 5'UTRs of a genome.

I have:

• Annotation file from MAKER (maker_annotation.gff3)
• Assembled genome (scaffold_genome_assembly.fasta)

So far, I've used GenomeTools to add intron features to my annotation file:

gt gff3 -addintrons maker_annotation.gff3 > intron_annotation.gff3

I created separate .gff3 files for the 5'UTRs and introns using grep:

cat intron_annotation_file.gff3 | grep five_prime_UTR > five_prime_UTR.gff3
cat intron_annotation_file.gff3 | grep intron > introns.gff3

With these files, I used bedtools to get a .gff3 of the intersection between introns and 5'UTRs:

bedtools intersect -wa -a introns.gff3 -b five_prime_UTR.gff3 > introns_five_prime.gff3

And used this last output to get the sequence from the genome:

bedtools getfasta -name -fo introns_five_seq.fasta -fi scaffold_genome_assembly.fasta -bed introns_five_prime.gff3

However, the sequences I get in introns_five_seq.fasta don't all start with GT and end in AG. Is there anything wrong with this process? I'm new to bioinformatics, any feedback is appreciated!

[husensofteng]

May be you want to remove the -wa option in bedtools intersect if you only want the intron fragments that are located within 5'UTRs as you have stated above otherwise you get the entire intron sequence.