DeepARG predict (version 1.0.4+galaxy0)

hide_db_build:

Input file:

DeepARG database:

Select model to use:

Molecular data type:

Minimum probability cutoff [Default: 0.8]:

Identity cutoff for sequence alignment [Default: 50]:

Evalue cutoff [Default: 1e-10]:

Alignment read overlap [Default: 0.8]:

Diamond, minimum number of alignments per entry [Default: 1000]:

Selection of the output files

Selection of the output files 0

DeepARG Predict is a computational tool designed to classify and annotate antibiotic resistance genes (ARGs) from nucleotide or protein sequences

It takes as input a fasta nucleotide or protein file containing short (SS model) or long (LS model) sequences

DeepARG output

DeepARG generates two main files: .ARG that contains the sequences with a probability sup or = --prob (0.8 default) and .potential.ARG with sequences containing a probability inf to --prob (0.8 default). The .potential.ARG file can still contain ARG-like sequences, howevere, it is necessary inspect its sequences

The output format for both files consists of the following fields:

ARG_NAME
QUERY_START
QUERY_END
QUERY_ID
PREDICTED_ARG_CLASS
BEST_HIT_FROM_DATABASE
PREDICTION_PROBABILITY
ALIGNMENT_BESTHIT_IDENTITY (%)
ALIGNMENT_BESTHIT_LENGTH
ALIGNMENT_BESTHIT_BITSCORE
ALIGNMENT_BESTHIT_EVALUE
COUNTS

If you want to annotate paired-end short read sequencing data use the DeepARG Short Reads tool