In Silico Tool for Predicting and Scanning Rheumatoid Arthritis-Inducing Peptides in an Antigen

Ritu TomerShipra JainPushpendra Singh GahlotNisha BajiyaGajendra PS Raghava^*

• Indraprastha Institute of Information Technology Delhi, Delhi, India

Rheumatoid arthritis (RA) is an autoimmune disorder in which the immune system mounts an abnormal response to self-antigens, leading to chronic inflammation and joint damage.Therefore, identifying antigenic regions in proteins that trigger RA is crucial for developing protein-based therapeutics. In this study, we developed models for predicting HLA class II binding RA-inducing peptides using a dataset comprising 291 experimentally confirmed RA-inducing peptides and 165 RA non-inducing peptides. Our initial analysis revealed that certain residues-such as glycine, proline, and tyrosine are significantly enriched in RAinducing peptides. While alignment based techniques like Basic Local Alignment Search Tool (BLAST) and Motif EmeRging and with Classes Identification (MERCI) offered high precision, they suffered from limited coverage. We developed machine and deep learningbased prediction models and obtained the highest performance (AUC = 0.75) using the XGBoost classifier on validation dataset. We also developed prediction methods using protein language models and achieved the highest performance (AUC = 0.72) using ProtBERT. Our ensemble model, which combines XGBoost and MERCI-derived motifs, achieved the best overall performance (AUC = 0.80; MCC = 0.45) on validation dataset. All models were rigorously evaluated using validation dataset that was not used during model training. This study will be valuable for assessing the risk of proteins used in probiotics, genetically modified foods, and protein-based therapeutics. Our most effective approach has been implemented in RAIpred, a web server and standalone software tool for predicting and scanning RA-inducing peptides (https://webs.iiitd.edu.in/raghava/raipred/).