Sismis

Machine learning-based secretion system annotation tool

🦇 Overview

Sismis (secretion system discovery tool; pronounced shish-mish) is a machine learning (ML)-based tool for detecting and classifying secretion systems in prokaryotic (meta)genomes.

🔍 Quickstart

To detect secretion systems in an assembled prokaryotic (meta)genome:

sismis run -g [fasta] -o [output directory] [options...]

For help/to view all options:

sismis -h

🔧 Installation

Sismis and its dependencies can be installed via pip:

pip install sismis

⚙️ Usage and options

Command structure

sismis run -g [fasta] -o [output directory] [options...]

Required arguments

    -g <file>, --genome <file>    a genomic file containing one or more
                                  sequences to use as input. Must be in
                                  one of the sequences format supported
                                  by Biopython.

🔖 Citation

If you found Sismis useful, please cite our preprint! 🤗

To cite Sismis and/or the Sismis Atlas:

Martin Larralde, Florian Albrecht, Josefin Blom, Johan Henriksson, Laura M Carroll. 2025. Scalable and interpretable secretion system annotation with Sismis. bioRxiv 2025.09.09.675188. doi: https://doi.org/10.1101/2025.09.09.675188.