Motivation In eukaryotes, proteins targeted for secretion contain a signal peptide, which allows them to proceed through the conventional ER/Golgi-dependent pathway. However, an important number of proteins lacking a signal peptide can be secreted through unconventional routes, including that mediated by exosomes. Currently, no method is available to predict protein secretion via exosomes. Results Here, we first assembled a dataset including the sequences of 2992 proteins secreted by exosomes and 2961 proteins that are not secreted by exosomes. Subsequently, we trained different random forests models on feature vectors derived from the sequences in this dataset. In tenfold cross-validation, the best model was trained on dipeptide composition, reaching an accuracy of 69.88% ± 2.08 and an area under the curve (AUC) of 0.76 ± 0.03. In an independent dataset, this model reached an accuracy of 75.73% and an AUC of 0.840. After these results, we developed ExoPred, a web-based tool that uses random forests to predict protein secretion by exosomes. Conclusion ExoPred is available for free public use at http://imath.med.ucm.es/exopred/. Datasets are available at http://imath.med.ucm.es/exopred/datasets/.
Prediction of linear B cell epitopes is of interest for the production of antigen-specific antibodies and the design of peptide-based vaccines. Here, we present BCEPS, a web server for predicting linear B cell epitopes tailored to select epitopes that are immunogenic and capable of inducing cross-reactive antibodies with native antigens. BCEPS implements various machine learning models trained on a dataset including 555 linearized conformational B cell epitopes that were mined from antibody–antigen protein structures. The best performing model, based on a support vector machine, reached an accuracy of 75.38% ± 5.02. In an independent dataset consisting of B cell epitopes retrieved from the Immune Epitope Database (IEDB), this model achieved an accuracy of 67.05%. In BCEPS, predicted epitopes can be ranked according to properties such as flexibility, accessibility and hydrophilicity, and with regard to immunogenicity, as judged by their predicted presentation by MHC II molecules. BCEPS also detects if predicted epitopes are located in ectodomains of membrane proteins and if they possess N-glycosylation sites hindering antibody recognition. Finally, we exemplified the use of BCEPS in the SARS-CoV-2 Spike protein, showing that it can identify B cell epitopes targeted by neutralizing antibodies.
Background We previously introduced PCPS (Proteasome Cleavage Prediction Server), a web-based tool to predict proteasome cleavage sites using n-grams. Here, we evaluated the ability of PCPS immunoproteasome cleavage model to discriminate CD8+ T cell epitopes. Results We first assembled an epitope dataset consisting of 844 unique virus-specific CD8+ T cell epitopes and their source proteins. We then analyzed cleavage predictions by PCPS immunoproteasome cleavage model on this dataset and compared them with those provided by a related method implemented by NetChop web server. PCPS was clearly superior to NetChop in term of sensitivity (0.89 vs. 0.79) but somewhat inferior with regard to specificity (0.55 vs. 0.60). Judging by the Mathew’s Correlation Coefficient, PCPS predictions were overall superior to those provided by NetChop (0.46 vs. 0.39). We next analyzed the power of C-terminal cleavage predictions provided by the same PCPS model to discriminate CD8+ T cell epitopes, finding that they could be discriminated from random peptides with an accuracy of 0.74. Following these results, we tuned the PCPS web server to predict CD8+ T cell epitopes and predicted the entire SARS-CoV-2 epitope space. Conclusions We report an improved version of PCPS named iPCPS for predicting proteasome cleavage sites and peptides with CD8+ T cell epitope features. iPCPS is available for free public use at https://imed.med.ucm.es/Tools/pcps/.
The outbreak of SARS-CoV-2 leading to the declaration of the COVID-19 global pandemic has led to the urgent development and deployment of several COVID-19 vaccines. Many of these new vaccines, including those based on mRNA and adenoviruses, are aimed to generate neutralizing antibodies against the spike glycoprotein, which is known to bind to the receptor angiotensin converting enzyme 2 (ACE2) in host cells via the receptor-binding domain (RBD). Antibodies binding to this domain can block the interaction with the receptor and prevent viral entry into the cells. Additionally, these vaccines can also induce spike-specific T cells which could contribute to providing protection against the virus. However, the emergence of new SARS-CoV-2 variants can impair the immunity generated by COVID-19 vaccines if mutations occur in cognate epitopes, precluding immune recognition. Here, we evaluated the chance of five SARS-CoV-2 variants of concern (VOCs), Alpha, Beta, Gamma, Delta and Omicron, to escape spike-specific immunity induced by vaccines. To that end, we examined the impact of the SARS-CoV-2 variant mutations on residues located on experimentally verified spike-specific epitopes, deposited at the Immune Epitope Database, that are targeted by neutralizing antibodies or recognized by T cells. We found about 300 of such B cell epitopes, which were largely overlapping, and could be grouped into 54 B cell epitope clusters sharing ≥ 7 residues. Most of the B cell epitope clusters map in the RBD domain (39 out of 54) and 20%, 50%, 37%, 44% and 57% of the total are mutated in SARS-CoV-2 Alpha, Beta, Gamma, Delta and Omicron variants, respectively. We also found 234 experimentally verified CD8 and CD4 T cell epitopes that were distributed evenly throughout the spike protein. Interestingly, in each SARS-CoV-2 VOC, over 87% and 79% of CD8 and CD4 T cell epitopes, respectively, are not mutated. These observations suggest that SARS-CoV-2 VOCs—particularly the Omicron variant—may be prone to escape spike-specific antibody immunity, but not cellular immunity, elicited by COVID-19 vaccines.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.