How to extract fasta sequences from assembled transcripts generated by Stringtie

Question

[seta]

Hi all,

I used STAR and stringtie for mapping reads to reference genome and assembly. As you know, the generated assembled transcripts by stringtie are in gtf format. Now, I want to have fasta sequence of assembled transcript. I used gffread, but all sequences had the same header! maybe it's not compatible with stringtie.

Could you please help me out to convert assembled transcripts by stringtie in gtf format to fasta format?

Thanks

[dukecomeback]

use gffread, you can find it in cufflink package

Answer 1

[zzqr]

The stringtie_merged.gtf file have seqname, start, end strand info. So, you can use R GRanges object and getSeq function from GenomicRanges and BSgenome packages to retrive sequences.

Answer 2

[Juke34]

I use agat_sp_extract_sequences.pl from AGAT.

agat_sp_extract_sequences.pl --cdna --gff input.gtf --fasta genome.fa -o output.fa

[Ishak]

thanks for your response

Answer 3

[lakhujanivijay]

You can also use bedtools getfasta to fetch sequences from GTF or BED files.

---

UPDATE

Here is the perfect solution

[spriyar]

I used this, but I run into the following error

"Error (GFaSeqGet): subsequence cannot be larger than 465 Error getting subseq for gene1 (465..1503)!"

Did you had any issues using gffread?

Thanks

[Cristina Zamora]

There is a Python script that fixes this error, you can follow A: gffread error when extracting transcript sequences from gtf, coordinates exceed

Answer 4

[DNAvinci]

Per, http://ccb.jhu.edu/software/stringtie/gff.shtml#gffread_ex

gffread \
-w assembled_transcripts.fa \
-g ref_genome.fa \
-E cov_refs.gtf \
./stringtie_output.gtf