How to get gene coding sequences based on GFF file?

Question

[biolab]

Hi everyone,

I have a draft genome fasta file and a GFF annotation file. The GFF file is like below.

9311_chr12      GLEAN   mRNA    17901210        17902763        0.90124 +       .       ID=9311_GLEAN_10008559;
9311_chr12      GLEAN   CDS     17901210        17901318        .       +       0       Parent=9311_GLEAN_10008559;
9311_chr12      GLEAN   CDS     17901418        17901486        .       +       2       Parent=9311_GLEAN_10008559;
9311_chr12      GLEAN   CDS     17901566        17901672        .       +       2       Parent=9311_GLEAN_10008559;
9311_chr12      GLEAN   CDS     17901722        17901755        .       +       0       Parent=9311_GLEAN_10008559;
9311_chr12      GLEAN   CDS     17902585        17902763        .       +       2       Parent=9311_GLEAN_10008559;
9311_chr04      GLEAN   mRNA    22207209        22208012        0.999282        -       .       ID=9311_GLEAN_10029041;
9311_chr04      GLEAN   CDS     22207209        22208012        .       -       0       Parent=9311_GLEAN_10029041;

My purpose is to get the gene coding sequences (without UTRs). I can filter the GFF file to include the CDS tracks only, but how to achieve the next step, that is to get the CDS sequences? Thank you very much!

[biolab]

Sorry for re-posting. I have found the solution on Biostars Extract Cds Fastas From A Gff Annotation + Reference Sequence

Thanks for your attention.

Answer 1

[PoGibas]

Reformat your GFF into CDS.bed and send it to bedtools getfasta.

# Not tested
awk '($3=="CDS") {OFS="\t"; print $1,$4,$5}' annotation.gff | 
    bedtools getfasta -fi genome.fa -bed - -fo CDS.fa

[biolab]

Thank you Pgibas, your comment is very helpful.