Biology subjects

Dorigatti, E.

Publications and source records attributed to Dorigatti, E..

4 recordsLinked to original sources

bioRxiv2023

Proteasomal cleavage prediction: state-of-the-art and future directions

Epitope vaccines are a promising approach for precision treatment of pathogens, cancer, autoimmune diseases, and allergies. Effectively designing such vaccines requires accurate proteasomal cleavage prediction to ensure that the epitopes included in the vaccine trigger an immune response. The performance of proteasomal cleavage predictors has been steadily improving over the past decades owing to increasing data availability and methodological advances. In this review, we summarize the current proteasomal cleavage prediction landscape and, in light of recent progress in the field of deep learning, develop and compare a wide range of recent architectures and techniques, including long short-term memory (LSTM), transformers, and convolutional neural networks (CNN), as well as four different denoising techniques. All open-source cleavage predictors re-trained on our dataset performed within two AUC percentage points. Our comprehensive deep learning architecture benchmark improved performance by 1.7 AUC percentage points, while closed-source predictors performed considerably worse. We found that a wide range of architectures and training regimes all result in very similar performance, suggesting that the specific modeling approach employed has a limited impact on predictive performance compared to the specifics of the dataset employed. We speculate that the noise and implicit nature of data acquisition techniques used for training proteasomal cleavage prediction models and the complexity of biological processes of the antigen processing pathway are the major limiting factors. While biological complexity can be tackled by more data and, to a lesser extent, better models, noise and randomness inherently limit the maximum achievable predictive performance. All our datasets and experiments are available at https://github.com/ziegler-ingo/cleavage_benchmark.

bioinformatics↗

bioRxiv2023

Predicting T Cell Receptor Functionality against Mutant Epitopes

Cancer cells or pathogens can escape recognition by T cell receptors (TCRs) through mutations of immunogenic epitopes. TCR cross-reactivity, i.e., recognition of multiple epitopes with sequence similarities, can be a factor to counteract such mutational escape. However, cross-reactivity of cell-based immunotherapies may also cause severe side effects when self-antigens are targeted. Therefore, the ability to predict the effect of mutations in the epitope sequence on T cell functionality in silico would greatly benefit the safety and effectiveness of newly-developed immunotherapies and vaccines. We here present "Predicting T cell Epitope-specific Activation against Mutant versions" (P-TEAM), a Random Forest-based model which predicts the effect of point mutations of an epitope on T cell functionality. We first trained and tested P-TEAM on a comprehensive dataset of 36 unique murine TCRs in response to systematic single-amino acid mutations of their target epitope (representing 5.472 unique TCR-epitope interactions). The model was able to classify T cell reactivities, corresponding to in vivo recruitment of T cells, and quantitatively predict T cell functionalities for unobserved single-point mutated altered peptide ligands (APLs), or even unseen TCRs, with consistently high performance. Further, we present an active learning framework to guide experimental design for assessing TCR functionality against novel epitopes, minimizing primary data acquisition costs. Finally, we applied P-TEAM to a novel dataset of 7 human TCRs reactive to the tumor neoantigen VPSVWRSSL. We observed a similarly robust performance for these human TCRs as for the murine TCRs recognizing SIINFEKL, thus providing evidence that our approach is applicable to therapeutically relevant TCRs as well as across species. Overall, P-TEAM provides an effective computational tool to study T cell responses against mutated epitopes.

bioinformatics↗

bioRxiv2020

Joint epitope selection and spacer design for string-of-beads vaccines

MotivationConceptually, epitope-based vaccine design poses two distinct problems: (1) selecting the best epitopes eliciting the strongest possible immune response, and (2) arranging and linking the selected epitopes through short spacer sequences to string-of-beads vaccines so as to increase the recovery likelihood of each epitope during antigen processing. Current state-of-the-art approaches solve this design problem sequentially. Consequently, such approaches are unable to capture the inter-dependencies between the two design steps, usually emphasizing theoretical immunogenicity over correct vaccine processing and resulting in vaccines with less effective immunogencity. ResultsIn this work, we present a computational approach based on linear programming that solves both design steps simultaneously, allowing to weigh the selection of a set of epitopes that have great immunogenic potential against their assembly into a string-of-beads construct that provides a high chance of recovery. We conducted Monte-Carlo cleavage simulations to show that, indeed, a fixed set of epitopes often cannot be assembled adequately, whereas selecting epitopes to accommodate proper cleavage requirements substantially improves their recovery probability and thus the effective immunogenicity, pathogen, and population coverage of the resulting vaccines by at least two fold. AvailabilityThe software and the data analyzed are available at https://github.com/SchubertLab/JessEV

bioinformatics↗

bioRxiv2019

Graph-Theoretical Formulation of the Generalized Epitope-based Vaccine Design Problem

Epitope-based vaccines have revolutionized vaccine research in the last decades. Due to their complex nature, bioinformatics plays a pivotal role in their development. However, existing algorithms address only specific parts of the design process or are unable to provide formal guarantees on the quality of the solution. Here we present a unifying formalism of the general epitope vaccine design problem that tackles all phases of the design process simultaneously and combines all prevalent design principles. We then demonstrate how to formulate the developed formalism as an integer linear program which guarantees optimality of the designs. This makes it possible to explore new regions of the vaccine design space, analyze the trade-offs between the design phases, and balance the many requirements of vaccines.

bioinformatics↗