Supplementary files: Identification and characterization of bacterial repeat-in-toxin adhesins using long-read genome analysis

This repository contains the AlphaFold models and modified third-party software supporting the manuscript Identification and characterization of bacterial repeat-in-toxin adhesins using long-read genome analysis. The primary codebase for LoRē, the workflow orchestrator used in this manuscript, can be found here: https://doi.org/10.5281/zenodo.21562775. 1. rtx_adhesin_models.tar.gz Contains the AlphaFold models and associated metadata for the RTX adhesins mentioned in the manuscript. All models were generated by Deep Mind's publicly available https://alphafoldserver.com/ _rtx_adhesin_models.csv: The master manifest of the adhesin models and associated metadata Model data: Each predicted structure includes: *.cif: The top-scoring predicted model *_summary_confidences.json: Global confidence metrics for the AlphaFold prediction *_job_request.json: The exact execution parameters used to generate the model Note: *_full_data.json containing full PAE confidences were not included because they were typically >100 MB each 2. FAL_prediction.tar.gz A redistributed copy of the FAL_prediction source code. We assume no authorship nor ownership of this pipeline and all rights belong to its authors: Monzon, V., Lafita, A., Bateman, A. Patched codebase: A minimally patched fork of the FAL_prediction pipeline made to work with Python 3.11 and pandas 2.3.2 T-ReksHPC.jar: A pre-built, distributable version of T-ReksHPC. This is included to allow the use of FAL_prediction without building anything from source.