Dataset Information

Combining Three-Dimensional Modeling with Artificial Intelligence to Increase Specificity and Precision in Peptide-MHC Binding Predictions.

ABSTRACT: The reliable prediction of the affinity of candidate peptides for the MHC is important for predicting their potential antigenicity and thus influences medical applications, such as decisions on their inclusion in T cell-based vaccines. In this study, we present a rapid, predictive computational approach that combines a popular, sequence-based artificial neural network method, NetMHCpan 4.0, with three-dimensional structural modeling. We find that the ensembles of bound peptide conformations generated by the programs MODELLER and Rosetta FlexPepDock are less variable in geometry for strong binders than for low-affinity peptides. In tests on 1271 peptide sequences for which the experimental dissociation constants of binding to the well-characterized murine MHC allele H-2D^b are known, by applying thresholds for geometric fluctuations the structure-based approach in a standalone manner drastically improves the statistical specificity, reducing the number of false positives. Furthermore, filtering candidates generated with NetMHCpan 4.0 with the structure-based predictor led to an increase in the positive predictive value (PPV) of the peptides correctly predicted to bind very strongly (i.e., K _d < 100 nM) from 40 to 52% (p = 0.027). The combined method also significantly improved the PPV when tested on five human alleles, including some with limited data for training. Overall, an average increase of 10% in the PPV was found over the standalone sequence-based method. The combined method should be useful in the rapid design of effective T cell-based vaccines.

SUBMITTER: Aranha MP

PROVIDER: S-EPMC7511449 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Novel therapeutic approaches for treating inherited retinal degenerations (IRDs) prompt a need to understand which patients with impaired vision have the anatomical potential to gain from participation in a clinical trial. We used supervised machine learning to predict foveal function from foveal structure in blue cone monochromacy (BCM), an X-linked congenital cone photoreceptor dysfunction secondary to mutations in the OPN1LW/OPN1MW gene cluster. BCM patients with either disease-associated large deletion or missense mutations were studied and results compared with those from subjects with other forms of IRD and various degrees of preserved central structure and function. A machine learning technique was used to associate foveal sensitivities and best-corrected visual acuities to foveal structure in IRD patients. Two random forest (RF) models trained on IRD data were applied to predict foveal function in BCM. A curve fitting method was also used and results compared with those of the RF models. The BCM and IRD patients had a comparable range of foveal structure. IRD patients had peak sensitivity at the fovea. Machine learning could successfully predict foveal sensitivity (FS) results from segmented or un-segmented optical coherence tomography (OCT) input. Application of machine learning predictions to BCM at the fovea showed differences between predicted and measured sensitivities, thereby defining treatment potential. The curve fitting method provided similar results. Given a measure of visual acuity (VA) and foveal outer nuclear layer thickness, the question of how many lines of acuity would represent the best efficacious result for each BCM patient could be answered. We propose that foveal vision improvement potential in BCM is predictable from retinal structure using machine learning and curve fitting approaches. This should allow estimates of maximal efficacy in patients being considered for clinical trials and also guide decisions about dosing.

Dataset Information

Combining Three-Dimensional Modeling with Artificial Intelligence to Increase Specificity and Precision in Peptide-MHC Binding Predictions.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets