DTSyn: a dual-transformer-based neural network to predict synergistic drug combinations

Abstract

Drug combination therapies are superior to monotherapy for cancer treatment in many ways. Identifying novel drug combinations by screening is challenging for the wet-lab experiments due to the time-consuming process of the enormous search space of possible drug pairs. Thus, computational methods have been developed to predict drug pairs with potential synergistic functions. Notwithstanding the success of current models, understanding the mechanism of drug synergy from a chemical–gene–tissue interaction perspective lacks study, hindering current algorithms from drug mechanism study. Here, we proposed a deep neural network model termed DTSyn (Dual Transformer encoder model for drug pair Synergy prediction) based on a multi-head attention mechanism to identify novel drug combinations. We designed a fine-granularity transformer encoder to capture chemical substructure–gene and gene–gene associations and a coarse-granularity transformer encoder to extract chemical–chemical and chemical–cell line interactions. DTSyn achieved the highest receiver operating characteristic area under the curve of 0.73, 0.78. 0.82 and 0.81 on four different cross-validation tasks, outperforming all competing methods. Further, DTSyn achieved the best True Positive Rate (TPR) over five independent data sets. The ablation study showed that both transformer encoder blocks contributed to the performance of DTSyn. In addition, DTSyn can extract interactions among chemicals and cell lines, representing the potential mechanisms of drug action. By leveraging the attention mechanism and pretrained gene embeddings, DTSyn shows improved interpretability ability. Thus, we envision our model as a valuable tool to prioritize synergistic drug pairs with chemical and cell line gene expression profile.

AttenSyn: An Attention-Based Deep Graph Neural Network for Anticancer Synergistic Drug Combination Prediction

“…Comparison with Existing Methods on Benchmark Datasets. To evaluate the effectiveness of the proposed AttenSyn, we compare it with several existing methods, including machine-learning-based methods (i.e., random forest (RF), Support Vector Machines (SVM), Multilayer Perceptron (MLP), Adaboost, and Elastic net) and deep-learning methods (i.e., DTSyn, 22 MR-GNN, 31 and DeepSynergy 14 ), by five fold cross validation on the benchmark datasets. The detailed comparative results are illustrated in Table 1, and the best results are shown in bold.…”

“…14 The synergy score of each drug pair was calculated by using the Combenefit tool. 26 According to a previous study, 22 we selected 10 as a threshold to classify the drug pair−cell-line triplets. The triplets with a synergistic score higher than 10 are recognized as positive, and those less than 0 are seen as negative.…”

“…To compare the performance of our proposed model with the state-of-the-art methods, we collected the drug combination dataset constructed by O’Neil et al as our benchmark dataset . The dataset contains 23,052 triplets, where each triplet comprises two drugs and a cancer cell line . There are 39 cancer cell lines and 38 unique drugs in the dataset, and these drugs are composed of 24 FDA-approved drugs and 14 experimental drugs .…”

“…25 The dataset contains 23,052 triplets, where each triplet comprises two drugs and a cancer cell line. 22 There are 39 cancer cell lines and 38 unique drugs in the dataset, and these drugs are composed of 24 FDA-approved drugs and 14 experimental drugs. 14 The synergy score of each drug pair was calculated by using the Combenefit tool.…”

“…Jin et al 21 proposed ComboNet, which employed a directional message-passing neural network (DMPNN) to learn a continuous representation of a molecule for identifying synergistic drug combinations for treating COVID-19. Additionally, Hu et al 22 developed a deep graph neural network model named DTSyn and used multihead attention mechanism to identify novel drug combinations. Although graph neural networks have demonstrated their ability to solve drug synergy prediction and achieved success to some extent in terms of performance improvement, there still exist some limitations.…”