xDeep-AcPEP: Deep Learning Method for Anticancer Peptide Activity Prediction Based on Convolutional Neural Network and Multitask Learning

Abstract

Cancer is one of the leading causes of death worldwide. Conventional cancer treatment relies on radiotherapy and chemotherapy, but both methods bring severe side effects to patients, as these therapies not only attack cancer cells but also damage normal cells. Anticancer peptides (ACPs) are a promising alternative as therapeutic agents that are efficient and selective against tumor cells. Here, we propose a deep learning method based on convolutional neural networks to predict biological activity (EC50, LC50, IC50, and LD50) against six tumor cells, including breast, colon, cervix, lung, skin, and prostate. We show that models derived with multitask learning achieve better performance than conventional single-task models. In repeated 5-fold cross validation using the CancerPPD data set, the best models with the applicability domain defined obtain an average mean squared error of 0.1758, Pearson's correlation coefficient of 0.8086, and Kendall's correlation coefficient of 0.6156. As a step toward model interpretability, we infer the contribution of each residue in the sequence to the predicted activity by means of feature importance weights derived from the convolutional layers of the model. The present method, referred to as xDeep-AcPEP, will help to identify effective ACPs in rational peptide design for therapeutic purposes. The data, script files for reproducing the experiments, and the final prediction models can be downloaded from http://github.com/chen709847237/xDeep-AcPEP. The web server to directly access this prediction method is at https://app.cbbio.online/acpep/home.

MLACP 2.0: An updated machine learning tool for anticancer peptide prediction

“…5 , we observed that six baseline models based on QSO encoding, and three baseline models based on 1OHE, CTDD, seq2vec, and CKSAAGP encoding respectively contributed two baseline models, four encodings (DDE, CTDC, KSC, and AAC) based their respective model contributed the most in the final MLACP 2.0 prediction. The importance of physicochemical properties has been highlighted in previous studies [49] , [50] . Our analysis also indicates that CKSAAGP is one of the influential features in MLACP 2.0 performance.…”

Modelling peptide–protein complexes: docking, simulations and machine learning

“…Along these lines, peptide-protein interaction predictors are also starting to emerge to predict which peptides will interact with a certain protein and give insight into the peptide residues involved in the interaction (Casadio et al, 2022;Lei et al, 2021). Furthermore, ML is also offering ways to identify peptide sequences which are likely to have high biological activity against a particular pathology (Wu et al, 2019) (e.g., anticancer (Chen et al, 2021) or antimicrobial (Dee, 2022;E. Y. Lee et al, 2017;Plisson et al, 2020) peptides).…”

Uncertainty quantification: Can we trust artificial intelligence in drug discovery?

“…Virtual screening (Berenger and Yamanishi, 2019) 2. Anticancer peptide activity prediction (Chen et al, 2021) 3. SARS-CoV 2 inhibitor prediction (Gawriljuk et al, 2021) 4.…”