Sofie Gielis

This paper introduces LIgO, a software suite that simulates adaptive immune receptor repertoires with immune information to aid in developing and benchmarking AIRR machine learning methods.

The paper analyzes the sequence distribution of epitope-specific TCR repertoires and shows how this affects machine learning model performance.

The paper develops computational models to identify WT1-specific TCRs in AML patients and finds higher TCR diversity in remission compared to relapsed patients.

The paper identifies challenges in predicting TCR-epitope interactions and introduces ImRex, a feature representation method that improves model performance for similar epitopes.

TCRex identifies epitope-specific T cell receptor sequences in repertoires using machine learning and statistical analysis, validated on vaccine and infectious disease datasets.

The paper evaluates unsupervised clustering methods for T-cell receptor sequences and finds distinct clusters that target the same epitopes.