ADEP: A Novel Approach Based on Discriminator-Enhanced Encoder-Decoder Architecture for Accurate Prediction of Adverse Effects in Polypharmacy

TL;DR

ADEP introduces a discriminator-enhanced encoder–decoder architecture to predict adverse effects in polypharmacy, addressing data sparsity and class imbalance through adversarial latent-space training. The model combines an autoencoder, classifier, and discriminator, optimized with a composite loss that balances reconstruction, classification, and adversarial objectives, with hyperparameters $\alpha=0.5$, $\beta=1$, $\gamma=1$. Evaluated on three large DDI benchmarks (DS1–DS3), ADEP outperforms 13 state-of-the-art methods across multiple metrics, and ablation studies confirm the critical role of the discriminator and neural classifier. A case study with DrugBank and TWOSIDES validation demonstrates practical applicability, highlighting ADEP’s potential to improve medication safety in polypharmacy settings. The work also discusses limitations, notably computational cost and data quality, and outlines avenues for broader deployment and further enhancement.

Abstract

Motivation: Unanticipated drug-drug interactions (DDIs) pose significant risks in polypharmacy, emphasizing the need for predictive methods. Recent advancements in computational techniques aim to address this challenge. Methods: We introduce ADEP, a novel approach integrating a discriminator and an encoder-decoder model to address data sparsity and enhance feature extraction. ADEP employs a three-part model, including multiple classification methods, to predict adverse effects in polypharmacy. Results: Evaluation on benchmark datasets shows ADEP outperforms well-known methods such as GGI-DDI, SSF-DDI, LSFC, DPSP, GNN-DDI, MSTE, MDF-SA-DDI, NNPS, DDIMDL, Random Forest, K-Nearest-Neighbor, Logistic Regression, and Decision Tree. Key metrics include Accuracy, AUROC, AUPRC, F-score, Recall, Precision, False Negatives, and False Positives. ADEP achieves more accurate predictions of adverse effects in polypharmacy. A case study with real-world data illustrates ADEP's practical application in identifying potential DDIs and preventing adverse effects. Conclusions: ADEP significantly advances the prediction of polypharmacy adverse effects, offering improved accuracy and reliability. Its innovative architecture enhances feature extraction from sparse medical data, improving medication safety and patient outcomes. Availability: Source code and datasets are available at https://github.com/m0hssn/ADEP.

Figures (1)

• Figure 1: The workflow of ADEP. a) Integration of drug feature vectors into a unified input vector through an encoder, facilitating comprehensive data representation for subsequent use by decoder, classifier, and discriminator components. b) Leveraging latent space for classification tasks and accurate input vector recreation by decoder. c) Utilization of latent space by classifier model for classification and prediction tasks. d) Generating samples from the latent distribution, training the model to distinguish between classes, and addressing data imbalance by identifying fake data closely resembling the original by the discriminator. e) Aggregate loss calculation: combining auxiliary losses from all three parts with coefficients to obtain final loss for backpropagation.