GFF3 Viewer and GFF3 to GenBank Converter
Read a GFF3 annotation file as a feature table, and turn it plus its FASTA into a GenBank record with an annotated map.
🔒 Nothing you paste is logged or stored on our servers
GFF3 is what genome browsers, Ensembl, NCBI's datasets downloads and every gene-prediction tool emit, and it is the format most plasmid editors cannot open. Paste one here to read it as a feature table: each feature's type, name, 1-based coordinates, strand, phase and parent, with percent-escapes decoded and a dot read as absent rather than as zero. The thing worth knowing about GFF3 is that it usually carries no sequence at all — it annotates a separate FASTA, joined on the first column — so this page tells you which of the two you have handed it, and takes the sequence separately when the file has none. Give it both and it writes a GenBank record you can open as an annotated map.
Paste a GFF3 file, or drop one anywhere on the page.
How to use the GFF3 Viewer & Converter tool
- 1Paste the GFF3 file, or drop it anywhere on the page.
- 2If the file has no embedded ##FASTA section, paste the DNA its coordinates refer to.
- 3Read the feature table, and open the converted GenBank record as an annotated map.
Frequently asked questions
Why does my GFF3 file have no sequence?
Because that is the normal case. GFF3 is a table of annotations that refers to a sequence held somewhere else — typically a FASTA file distributed alongside it — and the two are joined on the first column, the seqid. Some files embed the sequence after a ##FASTA line, and those are read whole. For the rest, paste the matching FASTA and the coordinates will be applied to it.
Will this read a GTF or GFF2 file?
No, and it says so rather than half-reading one. GTF and GFF2 use the same nine columns but a different attribute syntax (gene_id "X"; rather than ID=X), so the attributes would come back empty or wrong. Converting GTF to GFF3 first, with a tool such as gffread or AGAT, is the reliable route.
What happens to features that fall outside my sequence?
They are left out of the GenBank record and the omission is reported, with the furthest coordinate named. That combination almost always means the sequence pasted is not the one the GFF3 describes — a different assembly, or a different contig — which is worth catching before the map is drawn rather than after.