A Novel Autoencoder-Based Feature Selection Method for Drug-Target Interaction Prediction with Human-Interpretable Feature Weights
Drug-target interaction prediction provides important information that could be exploited for drug discovery, drug design, and drug repurposing. Chemogenomic approaches for predicting drug-target interaction assume that similar receptors bind to similar ligands. Capturing this similarity in so-calle...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2023-01-01
|
| Series: | Symmetry |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-8994/15/1/192 |
| _version_ | 1797436797929652224 |
|---|---|
| author | Gozde Ozsert Yigit Cesur Baransel |
| author_facet | Gozde Ozsert Yigit Cesur Baransel |
| author_sort | Gozde Ozsert Yigit |
| collection | DOAJ |
| description | Drug-target interaction prediction provides important information that could be exploited for drug discovery, drug design, and drug repurposing. Chemogenomic approaches for predicting drug-target interaction assume that similar receptors bind to similar ligands. Capturing this similarity in so-called “fingerprints” and combining the target and ligand fingerprints provide an efficient way to search for protein-ligand pairs that are more likely to interact. In this study, we constructed drug and target fingerprints by employing features extracted from the DrugBank. However, the number of extracted features is quite large, necessitating an effective feature selection mechanism since some features can be redundant or irrelevant to drug-target interaction prediction problems. Although such feature selection methods are readily available in the literature, usually they act as black boxes and do not provide any quantitative information about why a specific feature is preferred over another. To alleviate this lack of human interpretability, we proposed a novel feature selection method in which we used an autoencoder as a symmetric learning method and compared the proposed method to some popular feature selection algorithms, such as Kbest, Variance Threshold, and Decision Tree. The results of a detailed performance study, in which we trained six Multi-Layer Perceptron (MLP) Networks of different sizes and configurations for prediction, demonstrate that the proposed method yields superior results compared to the aforementioned methods. |
| first_indexed | 2024-03-09T11:07:49Z |
| format | Article |
| id | doaj.art-196498663dc8477b8781501a0c3470f2 |
| institution | Directory Open Access Journal |
| issn | 2073-8994 |
| language | English |
| last_indexed | 2024-03-09T11:07:49Z |
| publishDate | 2023-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Symmetry |
| spelling | doaj.art-196498663dc8477b8781501a0c3470f22023-12-01T00:53:15ZengMDPI AGSymmetry2073-89942023-01-0115119210.3390/sym15010192A Novel Autoencoder-Based Feature Selection Method for Drug-Target Interaction Prediction with Human-Interpretable Feature WeightsGozde Ozsert Yigit0Cesur Baransel1Department of Computer Engineering, Gaziantep University, 27410 Gaziantep, TurkeyDepartment of Computer Engineering, Gaziantep University, 27410 Gaziantep, TurkeyDrug-target interaction prediction provides important information that could be exploited for drug discovery, drug design, and drug repurposing. Chemogenomic approaches for predicting drug-target interaction assume that similar receptors bind to similar ligands. Capturing this similarity in so-called “fingerprints” and combining the target and ligand fingerprints provide an efficient way to search for protein-ligand pairs that are more likely to interact. In this study, we constructed drug and target fingerprints by employing features extracted from the DrugBank. However, the number of extracted features is quite large, necessitating an effective feature selection mechanism since some features can be redundant or irrelevant to drug-target interaction prediction problems. Although such feature selection methods are readily available in the literature, usually they act as black boxes and do not provide any quantitative information about why a specific feature is preferred over another. To alleviate this lack of human interpretability, we proposed a novel feature selection method in which we used an autoencoder as a symmetric learning method and compared the proposed method to some popular feature selection algorithms, such as Kbest, Variance Threshold, and Decision Tree. The results of a detailed performance study, in which we trained six Multi-Layer Perceptron (MLP) Networks of different sizes and configurations for prediction, demonstrate that the proposed method yields superior results compared to the aforementioned methods.https://www.mdpi.com/2073-8994/15/1/192drug-target interaction predictionautoencodermulti-layer perceptrondimensionality reductionfeature selectionDrugBank |
| spellingShingle | Gozde Ozsert Yigit Cesur Baransel A Novel Autoencoder-Based Feature Selection Method for Drug-Target Interaction Prediction with Human-Interpretable Feature Weights Symmetry drug-target interaction prediction autoencoder multi-layer perceptron dimensionality reduction feature selection DrugBank |
| title | A Novel Autoencoder-Based Feature Selection Method for Drug-Target Interaction Prediction with Human-Interpretable Feature Weights |
| title_full | A Novel Autoencoder-Based Feature Selection Method for Drug-Target Interaction Prediction with Human-Interpretable Feature Weights |
| title_fullStr | A Novel Autoencoder-Based Feature Selection Method for Drug-Target Interaction Prediction with Human-Interpretable Feature Weights |
| title_full_unstemmed | A Novel Autoencoder-Based Feature Selection Method for Drug-Target Interaction Prediction with Human-Interpretable Feature Weights |
| title_short | A Novel Autoencoder-Based Feature Selection Method for Drug-Target Interaction Prediction with Human-Interpretable Feature Weights |
| title_sort | novel autoencoder based feature selection method for drug target interaction prediction with human interpretable feature weights |
| topic | drug-target interaction prediction autoencoder multi-layer perceptron dimensionality reduction feature selection DrugBank |
| url | https://www.mdpi.com/2073-8994/15/1/192 |
| work_keys_str_mv | AT gozdeozsertyigit anovelautoencoderbasedfeatureselectionmethodfordrugtargetinteractionpredictionwithhumaninterpretablefeatureweights AT cesurbaransel anovelautoencoderbasedfeatureselectionmethodfordrugtargetinteractionpredictionwithhumaninterpretablefeatureweights AT gozdeozsertyigit novelautoencoderbasedfeatureselectionmethodfordrugtargetinteractionpredictionwithhumaninterpretablefeatureweights AT cesurbaransel novelautoencoderbasedfeatureselectionmethodfordrugtargetinteractionpredictionwithhumaninterpretablefeatureweights |