CRISPRimmunity predicts novel anti-CRISPR protein and achieves enhanced capability in analysis and prediction by applying ‘self-targeting’, ‘guilt by association’ and machine learning approaches, with the annotation of known Acrs.


Upload a prokaryotic genome sequence file and set corresponding parameters
Identify CRISPR-Cas system, ST event(s) and prophage region(s)
Predict anti-CRISPR protein (Acr) by ‘Homology search’ approach
Identify potential Acr by searching for homologs of 99 known Acrs using Diamond blastp with default parameters (see Parameter settings)
Predict anti-CRISPR protein (Acr) by 'Guilt by association' approach
Identify potential anti-CRISPR protein (Acr) by finding downstream Aca homologs
Predict anti-CRISPR protein (Acr) by 'CRISPR Self-Targeting prophage’ approach
 step1:Identify prophage(s) with self-targeting events
 step2:Detect HTH domain(s) in the prophage region(s)
 step3: Identify potential Acr by finding unknown functional and small size protein(s) neighboring the HTH domain protein (see Parameter settings)
 step4: If no HTH domain could be detected, then anti-CRISPR protein is predicted as the top ranking protein of the prophage region by AcRanker (machine learning)





Example 1: Predicted results for Staphylococcus schleiferi strain 5909-02, complete genome. One Type II CRISPR-Cas system was found with 16 spacers taregting prophages, plasmids and phages. Two AcrIIA13 proteins were detected in two prophages with protospacers, neighbored with an Aca.