Bug in negative strands of exons

[mpdunne]

There's a bug in the gff to gtf converter:

Deha2F_6        JGI     exon    544121  545199  .       -       .       gene_id "DEHA2F06138g"; transcript_id "DEHA2F06138g";
Deha2F_6        JGI     CDS     544121  545199  .       -       2       gene_id "DEHA2F06138g"; transcript_id "DEHA2F06138g";
Deha2F_6        JGI     stop_codon      544121  544123  .       -       0       gene_id "DEHA2F06138g"; transcript_id "DEHA2F06138g";
Deha2F_6        JGI     exon    545644  545872  .       -       .       gene_id "DEHA2F06138g"; transcript_id "DEHA2F06138g";
Deha2F_6        JGI     CDS     545644  545872  .       -       0       gene_id "DEHA2F06138g"; transcript_id "DEHA2F06138g";
Deha2F_6        JGI     start_codon     545870  545872  .       -       0       gene_id "DEHA2F06138g"; transcript_id "DEHA2F06138g";

converts to:

Deha2F_6        JGI     exon    544121  545199  .       -       .       gene_id "DEHA2F06138g"; transcript_id "DEHA2F06138g"; exon_number "1"; gene_name "DEHA2F06138g";
Deha2F_6        JGI     CDS     544121  544123  .       -       0       gene_id "DEHA2F06138g"; transcript_id "DEHA2F06138g"; exon_number "1"; gene_name "DEHA2F06138g";
Deha2F_6        JGI     start_codon     545870  545872  .       -       0       gene_id "DEHA2F06138g"; transcript_id "DEHA2F06138g"; exon_number "1"; gene_name "DEHA2F06138g";
Deha2F_6        JGI     exon    545644  545872  .       -       .       gene_id "DEHA2F06138g"; transcript_id "DEHA2F06138g"; exon_number "2"; gene_name "DEHA2F06138g";
Deha2F_6        JGI     CDS     544121  545199  .       -       2       gene_id "DEHA2F06138g"; transcript_id "DEHA2F06138g"; exon_number "2"; gene_name "DEHA2F06138g";
Deha2F_6        JGI     stop_codon      544121  544123  .       -       2       gene_id "DEHA2F06138g"; transcript_id "DEHA2F06138g"; exon_number "2"; gene_name "DEHA2F06138g";

Happens to negative strand genes only. CDS coordinates are completely wrong.

[vipints]

Just need to clarify one thing - the first file which you mentioned is not a GFF file. To me, the lines look like in GTF form. Primarily this program works for GFF3 to GTF conversion.