The script aims to fix the ORFs of gene models described in the gff file. By fixing it means replacing the original ORF (defined by the cds) when the longest predicted one within the mRNA is different. See the --model parameter for more details about the different cases. Currently the tool does not perform incomplete prediction (It always look for a start codon). It is consequently advised to not use the model5 except if you understand what you do. Several ouput files will be written if you specify an output. One will contain the gene not modified (intact), one with the gene models fixed (modified), one will both together (all).

SYNOPSIS -gff infile.gff --fasta genome.fa [ -o outfile ] --help


  • --gff

    Input GTF/GFF file.

  • -f, --fa or --fasta

    Imput fasta file.

  • --ct, --codon or --table

    Codon table to use. [default 1]

  • -m or --model

    Kind of ORF Model you want to fix. By default all are used. To select specific models writte e.g --model 1,4

    Model1 = The original ORF is part of the new ORF; the new ORF is longer

    Model2 = The original ORF and the new one are different; the new one is longer, they do not overlap each other.

    Model3 = The original ORF and the new one are different; the new one is longer, they overlap each other.

    Model4 = The new ORF is shorter due to the presence of stop codon in the original ORF.

    Model5 = The new ORF is shorter but the original ORF has not premature stop codon. The shorter predicted ORF can be due to the fact that the original ORF does not start by a start codon, while we force here the prediction to have a start codon. A ORF wihtout start can be the fact of an incomplete or fragmented ORF: annotation tool didn't predict the start because: * the start region is NNNN * the start region is XXXX * correct nucleotides but prediction tool did not annotate this part (e.g incomplete evidence in evidence-based prediction)

    Model6 = The ORF is same size but not correct frame (+1 or +2 bp gives a frame shift).

  • -s or --split

    This option is usefull for Model2. Indeed when the prediction is non overlapping the original cds, it is possible to split the gene into two different genes. By default we don't split it. We keep the longest. If you want to split it type: -s

  • -o , --output , --out or --outfile

    Output GFF file. If no output file is specified, the output will be written to STDOUT.

  • -v

    verbose mode. Default off. -v 1 minimum verbosity, -v 3 maximum verbosity

  • -c or --config

    String - Input agat config file. By default AGAT takes as input agat_config.yaml file from the working directory if any, otherwise it takes the orignal agat_config.yaml shipped with AGAT. To get the agat_config.yaml locally type: "agat config --expose". The --config option gives you the possibility to use your own AGAT config file (located elsewhere or named differently).

  • -h or --help

    Display this helpful text.