Chemical genomics language model toward reliable and explainable compound-protein interaction exploration.

BackgroundThe identification of compound-protein interactions (CPIs) is crucial for drug discovery and understanding mechanisms of action. Accurate CPI prediction can elucidate drug-target-disease interactions, aiding in the discovery of candidate compounds and effective synergistic drugs, particularly from traditional Chinese medicine (TCM). Existing in silico methods face challenges in prediction accuracy and generalization due to compound and target diversity and the lack of largescale interaction datasets and negative datasets for model learning.MethodsTo address these issues, we developed a computational model for CPI prediction by integrating the constructed large-scale bioactivity benchmark dataset with a deep learning (DL) algorithm. To verify the accuracy of our CPI model, we applied it to predict the targets of compounds in TCM. An herb pair of Astragalus membranaceus and Hedyotis diffusaas was used as a model, and the active compounds in this herb pair were collected from various public databases and the literature. The complete targets of these active compounds were predicted by the CPI model, resulting in an expanded target dataset. This dataset was next used for the prediction of synergistic antitumor compound combinations. The predicted multi-compound combinations were subsequently examined through in vitro cellular experiments.ResultsOur CPI model demonstrated superior performance over other machine learning models, achieving an area under the Receiver Operating Characteristic curve (AUROC) of 0.98, an area under the precision-recall curve (AUPR) of 0.98, and an accuracy (ACC) of 93.31% on the test set. The model’s generalization capability and applicability were further confirmed using external databases. Utilizing this model, we predicted the targets of compounds in the herb pair of Astragalus membranaceus and Hedyotis diffusaas, yielding an expanded target dataset. Then, we integrated this expanded target dataset to predict effective drug combinations using our drug synergy prediction model DeepMDS. Experimental assay on breast cancer cell line MDA-MB-231 proved the efficacy of the best predicted multi-compound combinations: Combination I (Epicatechin, Ursolic acid, Quercetin, Aesculetin and Astragaloside IV) exhibited a half-maximal inhibitory concentration (IC50) value of 19.41 μM, and a combination index (CI) value of 0.682; and Combination II (Epicatechin, Ursolic acid, Quercetin, Vanillic acid and Astragaloside IV) displayed a IC50 value of 23.83 μM and a CI value of 0.805. These results validated the ability of our model to make accurate predictions for novel CPI data outside the training dataset and evaluated the reliability of the predictions, showing good applicability potential in drug discovery and in the elucidation of the bioactive compounds in TCM.ConclusionOur CPI prediction model can serve as a useful tool for accurately identifying potential CPI for a wide range of proteins, and is expected to facilitate drug research, repurposing and support the understanding of TCM.

Identifying compound-protein interactions (CPIs) is critical in drug discovery, as accurate prediction of CPIs can remarkably reduce the time and cost of new drug development. The rapid growth of existing biological knowledge has opened up possibilities for leveraging known biological knowledge to predict unknown CPIs. However, existing CPI prediction models still fall short of meeting the needs of practical drug discovery applications. A novel parallel graph convolutional network model for CPI prediction (ParaCPI) is proposed in this study. This model constructs feature representation of compounds using a unique approach to predict unknown CPIs from known CPI data more effectively. Experiments are conducted on five public datasets, and the results are compared with current state-of-the-art (SOTA) models under three different experimental settings to evaluate the model's performance. In the three cold-start settings, ParaCPI achieves an average performance gain of 26.75%, 23.84%, and 14.68% in terms of area under the curve compared with the other SOTA models. In addition, the results of the experiments in the case study show ParaCPI's superior ability to predict unknown CPIs based on known data, with higher accuracy and stronger generalization compared with the SOTA models. Researchers can leverage ParaCPI to accelerate the drug discovery process.

Accurate prediction of compound-protein interaction (CPI) plays an important role in drug discovery and bioinformatics. While graph neural networks have attracted increasing interest for molecular graph representation in CPI prediction, they perform poorly in capturing global structural features of compounds. Moreover, the integration of protein physicochemical properties remains limited in current CPI models. To address these limitations, we propose the TFT-CPI model, a novel feature-enhanced CPI prediction model based on topology and fractal theory. Topological data analysis can effectively capture molecular global structural features and sequence fractal dimension is used to extract hydrophilicity information. Specifically, we use graph convolutional networks combined with persistence landscapes to extract molecular features and use convolutional neural networks combined with protein sequence fractal dimension for protein feature extraction. The model also incorporates a symmetric Transformer architecture to improve prediction performance. We conducted the experiments on seven different benchmark datasets, including the Human and C.elegans datasets with varying positive-to-negative sample ratios, as well as a label reversal dataset, GPCR. TFT-CPI model achieved an average AUC improvement of 0.4% on Human and C.elegans datasets compared to the state-of-the-art models and attained the highest AUC and PRC on GPCR dataset compared to the baselines, reaching 86.4% and 87.6%, respectively. These results highlight its superior performance and strong generalization capabilities. Furthermore, TFT-CPI model exhibits promising performance in out-of-distribution and cross-dataset scenarios, further demonstrating its effectiveness across diverse CPI prediction tasks.

FOTF-CPI: A compound-protein interaction prediction transformer based on the fusion of optimal transport fragments

Compound–protein interaction prediction by deep learning: Databases, descriptors and models

Computational prediction of compound-protein interactions (CPIs) is of great importance for drug design as the first step in in-silico screening. We previously proposed chemical genomics-based virtual screening (CGBVS), which predicts CPIs by using a support vector machine (SVM). However, the CGBVS has problems when training using more than a million datasets of CPIs since SVMs require an exponential increase in the calculation time and computer memory. To solve this problem, we propose the CGBVS-DNN, in which we use deep neural networks, a kind of deep learning technique, instead of the SVM. Deep learning does not require learning all input data at once because the network can be trained with small mini-batches. Experimental results show that the CGBVS-DNN outperformed the original CGBVS with a quarter million CPIs. Results of cross-validation show that the accuracy of the CGBVS-DNN reaches up to 98.2 % (σ<0.01) with 4 million CPIs.

Heterogeneous sampled subgraph neural networks with knowledge distillation to enhance double-blind compound-protein interaction prediction

In silico prediction of compound-protein interactions from heterogeneous biological data is critical in the process of drug development. In this chapter the authors review several supervised machine learning methods to predict unknown compound-protein interactions from chemical structure and genomic sequence information simultaneously. The authors review several kernel-based algorithms from two different viewpoints: binary classification and dimension reduction. In the results, they demonstrate the usefulness of the methods on the prediction of drug-target interactions and ligand-protein interactions from chemical structure data and genomic sequence data.

Identifying compound-protein interactions (CPIs) is crucial for drug discovery. Since experimentally validating CPIs is often time-consuming and costly, computational approaches are expected to facilitate the process. Rapid growths of available CPI databases have accelerated the development of many machine-learning methods for CPI predictions. However, their performance, particularly their generalizability against external data, often suffers from a data imbalance attributed to the lack of experimentally validated inactive (negative) samples. In this study, we developed a self-training method for augmenting both credible and informative negative samples to improve the performance of models impaired by data imbalances. The constructed model demonstrated higher performance than those constructed with other conventional methods for solving data imbalances, and the improvement was prominent for external datasets. Moreover, examination of the prediction score thresholds for pseudo-labeling during self-training revealed that augmenting the samples with ambiguous prediction scores is beneficial for constructing a model with high generalizability. The present study provides guidelines for improving CPI predictions on real-world data, thus facilitating drug discovery.

Co-guided Dual-channel Graph Neural Networks for the prediction of compound–protein interaction

Predicting binding affinities between small molecules and the protein target is at the core of computational drug screening and drug target identification. Deep learning-based approaches have recently been adapted to predict binding affinities and they claim to achieve high prediction accuracy in their tests; we show that these approaches do not generalize, that is, they fail to predict interactions between unknown proteins and unknown small molecules. To address these shortcomings, we develop a new compound-protein interaction predictor, Yuel, which predicts compound-protein interactions with a higher generalizability than the existing methods. Upon comprehensive tests on various data sets, we find that out of all the deep-learning approaches surveyed, Yuel manifests the best ability to predict interactions between unknown compounds and unknown proteins.

Motivation: Computational prediction of compound–protein interactions (CPIs) is of great importance for drug design and development, as genome-scale experimental validation of CPIs is not only time-consuming but also prohibitively expensive. With the availability of an increasing number of validated interactions, the performance of computational prediction approaches is severely impended by the lack of reliable negative CPI samples. A systematic method of screening reliable negative sample becomes critical to improving the performance of in silico prediction methods.Results: This article aims at building up a set of highly credible negative samples of CPIs via an in silico screening method. As most existing computational models assume that similar compounds are likely to interact with similar target proteins and achieve remarkable performance, it is rational to identify potential negative samples based on the converse negative proposition that the proteins dissimilar to every known/predicted target of a compound are not much likely to be targeted by the compound and vice versa. We integrated various resources, including chemical structures, chemical expression profiles and side effects of compounds, amino acid sequences, protein–protein interaction network and functional annotations of proteins, into a systematic screening framework. We first tested the screened negative samples on six classical classifiers, and all these classifiers achieved remarkably higher performance on our negative samples than on randomly generated negative samples for both human and Caenorhabditis elegans. We then verified the negative samples on three existing prediction models, including bipartite local model, Gaussian kernel profile and Bayesian matrix factorization, and found that the performances of these models are also significantly improved on the screened negative samples. Moreover, we validated the screened negative samples on a drug bioactivity dataset. Finally, we derived two sets of new interactions by training an support vector machine classifier on the positive interactions annotated in DrugBank and our screened negative interactions. The screened negative samples and the predicted interactions provide the research community with a useful resource for identifying new drug targets and a helpful supplement to the current curated compound–protein databases.Availability: Supplementary files are available at: http://admis.fudan.edu.cn/negative-cpi/. Contact: sgzhou@fudan.edu.cnSupplementary Information: Supplementary data are available at Bioinformatics online.

DeepCPI: A Deep Learning-based Framework for Large-scale in silico Drug Screening

EAAT2 is a high affinity, Na+-dependent glutamate transporter with predominant astroglial localization. It accounts for the clearance of the bulk of glutamate released at central nervous system synapses and therefore has a crucial role in shaping glutamatergic neurotransmission and limiting excitotoxicity. Caspase-3 activation and impairment in expression and activity of EAAT2 are two distinct molecular mechanisms occurring in human amyotrophic lateral sclerosis (ALS) and in the transgenic rodent model of the disease. Excitotoxicity caused by down-regulation of EAAT2 is thought to be a contributing factor to motor neuron death in ALS. In this study, we report the novel evidence that caspase-3 cleaves EAAT2 at a unique site located in the cytosolic C-terminal domain of the transporter, a finding that links excitotoxicity and activation of caspase-3 as converging mechanisms in the pathogenesis of ALS. Caspase-3 cleavage of EAAT2 leads to a drastic and selective inhibition of this transporter. Heterologous expression of mutant SOD1 proteins linked to the familial form of ALS leads to inhibition of EAAT2 through a mechanism that largely involves activation of caspase-3 and cleavage of the transporter. In addition, we found evidence in spinal cord homogenates of mutant SOD1 ALS mice of a truncated form of EAAT2, likely deriving from caspase-3-mediated proteolytic cleavage, which appeared concurrently to the loss of EAAT2 immunoreactivity and to increased expression of activated caspase-3. Taken together, our findings suggest that caspase-3 cleavage of EAAT2 is one mechanism responsible for the impairment of glutamate uptake in mutant SOD1-linked ALS.

Objective To study the effect of intestinal bacteria on motor ability of amyotrophic lateral sclerosis (ALS) mice models and its mechanism. Methods Twenty wild type C57BL/6J mice (WT group) and 20 SOD1-G93A transgenic ALS mice (ALS group) were selected as the research subjects. (1) Ten mice in both WT group and ALS group were selected, respectively; 5 mice in each group were fed in SPF environment, and the remaining 5 mice were fed in aseptic environment; they were defined as WT+SPF group, WT+aseptic group, ALS+SPF group and ALS+aseptic group. (2) Ten mice in WT group and ALS group were fed in sterile environment; 5 mice in each group were transplanted with fecal bacteria, and the remaining 5 mice were not interfered; they were defined as WT+transplantation group, WT+non-transplantation group, ALS+transplantation group and ALS+non transplantation group. The grip strength of mice was measured by grip force meter, the motor coordination ability of mice was tested by roller treadmill and rotating rod test, the number of motor neurons in the anterior horn of spinal cord was measured by Nissl staining, the expression of microglia activation marker ionic calcium junction protein (IBA-1) in spinal cord tissues was detected by immunohistochemical staining, and the expressions of tumor necrosis factor (TNF)-α and interleukin (IL)-6 in spinal cord tissues were detected by Western blotting; the β-N-methylamino-L-alanine (BMAA) expression was detected by high performance liquid chromatography-tandem mass spectrometry. Results (1) The grip strength, drop latency and drop time of ALS+aseptic mice were significantly higher than those of ALS+SPF mice, the number of Nissl-stained positive cells was significantly larger than that of ALS+SPF mice, the number of IBA-1 positive cells was significantly smaller than that of ALS+SPF mice, the levels of TNF-α and IL-6 protein expressions and BMAA concentration were statistically lower than those of ALS+SPF mice (P<0.05). (2) The grip strength, drop latency and drop time of ALS+transplantion mice were significantly lower than those of ALS+non-transplantation mice, the number of Nissl-stained positive cells was significantly smaller than that of ALS+non-transplantation mice, the number of IBA-1 positive cells was significantly larger than that of ALS+non-transplantation mice, the TNF-α and IL-6 protein expressions and BMAA concentration were significantly higher than those of ALS+non-transplantation mice (P<0.05). Conclusion Imbalance of intestinal bacteria homeostasis can decrease the motor ability of ALS mice, which is related to the activation of microglia. Key words: Amyotrophic lateral sclerosis; Intestinal bacteria; Motor ability; Motor neuron; Microglia