Graph regularized non-negative matrix factorization with $$L_{2,1}$$ L 2 , 1 norm regularization terms for drug–target interactions prediction
Abstract Background Identifying drug–target interactions (DTIs) plays a key role in drug development. Traditional wet experiments to identify DTIs are costly and time consuming. Effective computational methods to predict DTIs are useful to speed up the process of drug discovery. A variety of non-neg...
| Main Authors: | , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
BMC
2023-10-01
|
| Series: | BMC Bioinformatics |
| Subjects: | |
| Online Access: | https://doi.org/10.1186/s12859-023-05496-6 |
| _version_ | 1797451290680229888 |
|---|---|
| author | Junjun Zhang Minzhu Xie |
| author_facet | Junjun Zhang Minzhu Xie |
| author_sort | Junjun Zhang |
| collection | DOAJ |
| description | Abstract Background Identifying drug–target interactions (DTIs) plays a key role in drug development. Traditional wet experiments to identify DTIs are costly and time consuming. Effective computational methods to predict DTIs are useful to speed up the process of drug discovery. A variety of non-negativity matrix factorization based methods are proposed to predict DTIs, but most of them overlooked the sparsity of feature matrices and the convergence of adopted matrix factorization algorithms, therefore their performances can be further improved. Results In order to predict DTIs more accurately, we propose a novel method iPALM-DLMF. iPALM-DLMF models DTIs prediction as a problem of non-negative matrix factorization with graph dual regularization terms and $$L_{2,1}$$ L 2 , 1 norm regularization terms. The graph dual regularization terms are used to integrate the information from the drug similarity matrix and the target similarity matrix, and $$L_{2,1}$$ L 2 , 1 norm regularization terms are used to ensure the sparsity of the feature matrices obtained by non-negative matrix factorization. To solve the model, iPALM-DLMF adopts non-negative double singular value decomposition to initialize the nonnegative matrix factorization, and an inertial Proximal Alternating Linearized Minimization iterating process, which has been proved to converge to a KKT point, to obtain the final result of the matrix factorization. Extensive experimental results show that iPALM-DLMF has better performance than other state-of-the-art methods. In case studies, in 50 highest-scoring proteins targeted by the drug gabapentin predicted by iPALM-DLMF, 46 have been validated, and in 50 highest-scoring drugs targeting prostaglandin-endoperoxide synthase 2 predicted by iPALM-DLMF, 47 have been validated. |
| first_indexed | 2024-03-09T14:52:36Z |
| format | Article |
| id | doaj.art-bc857e5370954f2c8f7717301677d9c3 |
| institution | Directory Open Access Journal |
| issn | 1471-2105 |
| language | English |
| last_indexed | 2024-03-09T14:52:36Z |
| publishDate | 2023-10-01 |
| publisher | BMC |
| record_format | Article |
| series | BMC Bioinformatics |
| spelling | doaj.art-bc857e5370954f2c8f7717301677d9c32023-11-26T14:23:06ZengBMCBMC Bioinformatics1471-21052023-10-0124112410.1186/s12859-023-05496-6Graph regularized non-negative matrix factorization with $$L_{2,1}$$ L 2 , 1 norm regularization terms for drug–target interactions predictionJunjun Zhang0Minzhu Xie1Key Laboratory of Computing and Stochastic Mathematics(LCSM) (Ministry of Education), School of Mathematics and Statistics, Hunan Normal UniversityKey Laboratory of Computing and Stochastic Mathematics(LCSM) (Ministry of Education), School of Mathematics and Statistics, Hunan Normal UniversityAbstract Background Identifying drug–target interactions (DTIs) plays a key role in drug development. Traditional wet experiments to identify DTIs are costly and time consuming. Effective computational methods to predict DTIs are useful to speed up the process of drug discovery. A variety of non-negativity matrix factorization based methods are proposed to predict DTIs, but most of them overlooked the sparsity of feature matrices and the convergence of adopted matrix factorization algorithms, therefore their performances can be further improved. Results In order to predict DTIs more accurately, we propose a novel method iPALM-DLMF. iPALM-DLMF models DTIs prediction as a problem of non-negative matrix factorization with graph dual regularization terms and $$L_{2,1}$$ L 2 , 1 norm regularization terms. The graph dual regularization terms are used to integrate the information from the drug similarity matrix and the target similarity matrix, and $$L_{2,1}$$ L 2 , 1 norm regularization terms are used to ensure the sparsity of the feature matrices obtained by non-negative matrix factorization. To solve the model, iPALM-DLMF adopts non-negative double singular value decomposition to initialize the nonnegative matrix factorization, and an inertial Proximal Alternating Linearized Minimization iterating process, which has been proved to converge to a KKT point, to obtain the final result of the matrix factorization. Extensive experimental results show that iPALM-DLMF has better performance than other state-of-the-art methods. In case studies, in 50 highest-scoring proteins targeted by the drug gabapentin predicted by iPALM-DLMF, 46 have been validated, and in 50 highest-scoring drugs targeting prostaglandin-endoperoxide synthase 2 predicted by iPALM-DLMF, 47 have been validated.https://doi.org/10.1186/s12859-023-05496-6Drug–target interactions$$L_{2,1}$$ L 2 , 1 normInertial proximal alternating linearized minimization |
| spellingShingle | Junjun Zhang Minzhu Xie Graph regularized non-negative matrix factorization with $$L_{2,1}$$ L 2 , 1 norm regularization terms for drug–target interactions prediction BMC Bioinformatics Drug–target interactions $$L_{2,1}$$ L 2 , 1 norm Inertial proximal alternating linearized minimization |
| title | Graph regularized non-negative matrix factorization with $$L_{2,1}$$ L 2 , 1 norm regularization terms for drug–target interactions prediction |
| title_full | Graph regularized non-negative matrix factorization with $$L_{2,1}$$ L 2 , 1 norm regularization terms for drug–target interactions prediction |
| title_fullStr | Graph regularized non-negative matrix factorization with $$L_{2,1}$$ L 2 , 1 norm regularization terms for drug–target interactions prediction |
| title_full_unstemmed | Graph regularized non-negative matrix factorization with $$L_{2,1}$$ L 2 , 1 norm regularization terms for drug–target interactions prediction |
| title_short | Graph regularized non-negative matrix factorization with $$L_{2,1}$$ L 2 , 1 norm regularization terms for drug–target interactions prediction |
| title_sort | graph regularized non negative matrix factorization with l 2 1 l 2 1 norm regularization terms for drug target interactions prediction |
| topic | Drug–target interactions $$L_{2,1}$$ L 2 , 1 norm Inertial proximal alternating linearized minimization |
| url | https://doi.org/10.1186/s12859-023-05496-6 |
| work_keys_str_mv | AT junjunzhang graphregularizednonnegativematrixfactorizationwithl21l21normregularizationtermsfordrugtargetinteractionsprediction AT minzhuxie graphregularizednonnegativematrixfactorizationwithl21l21normregularizationtermsfordrugtargetinteractionsprediction |