PrediTALE: A novel model learned from quantitative data allows for new perspectives on TALE targeting

Area of Science:

• Plant pathology
• Molecular biology
• Bioinformatics

Background:

• Plant-pathogenic Xanthomonas bacteria utilize transcription activator-like effectors (TALEs) to manipulate host gene expression and promote virulence.
• TALEs possess a modular DNA-binding domain with repeat-variable diresidues (RVDs) that determine DNA target specificity.
• Understanding TALE-DNA interactions is crucial for deciphering bacterial virulence mechanisms.

Purpose of the Study:

• To develop and validate a novel computational approach for predicting TALE target genes.
• To improve the accuracy of TALE target prediction by incorporating recent findings in TALE-DNA binding.
• To identify novel virulence targets of TALEs, including those associated with orphan TALEs.

Main Methods:

• A novel computational model, PrediTALE, was developed to predict TALE target DNA sequences.
• The model incorporates TALE targeting features, including aberrant repeat lengths and flexible strand orientation.
• PrediTALE was benchmarked using RNA-sequencing data from rice plants infected with Xanthomonas.

Main Results:

• PrediTALE demonstrated improved prediction performance compared to existing methods.
• The tool successfully predicted several novel putative virulence targets for TALEs.
• The study identified 'orphan TALEs' with no predicted targets, suggesting issues with current gene annotations.

Conclusions:

• PrediTALE offers a more accurate method for predicting TALE virulence targets.
• Genome-wide scans combined with RNA-seq data can identify TALE targets independent of gene annotations.
• This approach advances our understanding of bacterial virulence and host-pathogen interactions.