Genome annotation · Software
Prodigal
Protein-coding gene prediction software for bacterial, archaeal, and metagenomic DNA sequences.
What Prodigal is
Prodigal is a gene prediction tool designed to locate protein-coding genes in prokaryotic organisms, including bacteria and archaea. Developed by Doug Hyatt, it provides automated annotation for finished genomes, draft assemblies, and metagenomic sequences.
Annotating newly sequenced prokaryotes often requires custom training models to accurately identify coding regions and start sites. Prodigal addresses this by applying an unsupervised machine learning algorithm that learns genome properties directly from the input sequence without requiring external training data.
How it works
Prodigal processes raw sequence data by automatically learning sequence characteristics directly from the genome. It determines coding statistics, start codon usage, and ribosome binding site (RBS) motif preferences without prior training.
Using these learned properties, it predicts protein-coding gene coordinates and identifies translation initiation sites. For each candidate start site, it can output confidence scores and RBS motif information. The algorithm also provides options to dictate whether genes should span across runs of unknown bases (N's) and how partial genes at contig boundaries should be handled.
Key facts
- Version
- 2.6.3
- Category
- Genome annotation
- Target organisms
- Bacteria and Archaea
- Supported inputs
- Finished genomes, draft genomes, and metagenomes
- Output formats
- GFF3, Genbank, Sequin table, and FASTA protein translations
- Learning approach
- Unsupervised machine learning
What you can use it for
- Prokaryotic genome annotation
- Predict protein-coding genes and generate translated protein sequences from finished or draft bacterial and archaeal genomes.
- Metagenomic gene discovery
- Identify coding sequences within uncultured community DNA using the dedicated metagenomic analysis mode.
- Translation initiation site mapping
- Locate and score candidate start sites across a genome, obtaining associated RBS motifs and confidence values.
- Draft contig analysis
- Identify coding regions on fragmented contigs with custom settings for partial genes at contig edges and sequence gaps.
Run it here
Prodigal is preinstalled on the machine behind every research conversation. Ask the agent in plain language, or give it the exact command — it runs it, reads the output and explains it.
Strengths and limits
Good at
- Operates as an unsupervised algorithm that learns coding properties without requiring pre-existing training sets.
- Executes rapidly, analysing an entire Escherichia coli K-12 genome in approximately 10 seconds.
- Supports multiple input data types, including finished genomes, draft assemblies, and metagenomes.
- Provides detailed start-site predictions, including confidence scores and ribosomal binding site motifs.
Watch out for
- Restricted to prokaryotic organisms and does not support eukaryotic gene prediction.
- Requires user-defined parameter choices to determine how genes crossing gaps of N's or truncated contig edges are processed.
Questions
Can I run Prodigal online without installing it?
Prodigal is preinstalled on the machine behind every research conversation. Ask the agent in plain language, or give it the exact command — it runs it, reads the output and explains it.
What licence is Prodigal under?
Prodigal is published under its authors' own licence; the project page at https://github.com/hyattpd/Prodigal states it.
Does Prodigal require pre-compiled training models or datasets?
No. Prodigal is an unsupervised machine learning algorithm that automatically learns coding statistics, start codon usage, and RBS motifs directly from the sequence itself.
Can Prodigal run on fragmented assemblies or metagenomic data?
Yes. It accommodates draft genomes, metagenomes, and finished genomes, and includes a specific metagenomic procedure.
What annotation file formats can Prodigal generate?
Prodigal outputs gene predictions in GFF3, Genbank, and Sequin table formats, and can also write translated protein sequences to a separate file.
How does the tool handle incomplete genes at contig boundaries?
Users can specify options that control how partial genes situated at the edges of contigs are treated.
Written from Prodigal's own README and project page. Statements citing a figure that does not appear there are removed before publication. Licence and running details come from our registry.