Protein Docking Model Evaluation by Graph Neural Networks
Physical interactions of proteins play key functional roles in many important cellular processes. To understand molecular mechanisms of such functions, it is crucial to determine the structure of protein complexes. To complement experimental approaches, which usually take a considerable amount of ti...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2021-05-01
|
| Series: | Frontiers in Molecular Biosciences |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fmolb.2021.647915/full |
| _version_ | 1818647591870529536 |
|---|---|
| author | Xiao Wang Sean T. Flannery Daisuke Kihara Daisuke Kihara |
| author_facet | Xiao Wang Sean T. Flannery Daisuke Kihara Daisuke Kihara |
| author_sort | Xiao Wang |
| collection | DOAJ |
| description | Physical interactions of proteins play key functional roles in many important cellular processes. To understand molecular mechanisms of such functions, it is crucial to determine the structure of protein complexes. To complement experimental approaches, which usually take a considerable amount of time and resources, various computational methods have been developed for predicting the structures of protein complexes. In computational modeling, one of the challenges is to identify near-native structures from a large pool of generated models. Here, we developed a deep learning–based approach named Graph Neural Network–based DOcking decoy eValuation scorE (GNN-DOVE). To evaluate a protein docking model, GNN-DOVE extracts the interface area and represents it as a graph. The chemical properties of atoms and the inter-atom distances are used as features of nodes and edges in the graph, respectively. GNN-DOVE was trained, validated, and tested on docking models in the Dockground database and further tested on a combined dataset of Dockground and ZDOCK benchmark as well as a CAPRI scoring dataset. GNN-DOVE performed better than existing methods, including DOVE, which is our previous development that uses a convolutional neural network on voxelized structure models. |
| first_indexed | 2024-12-17T01:04:58Z |
| format | Article |
| id | doaj.art-6a8c3c5c0e274c16b38005b4c3851f24 |
| institution | Directory Open Access Journal |
| issn | 2296-889X |
| language | English |
| last_indexed | 2024-12-17T01:04:58Z |
| publishDate | 2021-05-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Molecular Biosciences |
| spelling | doaj.art-6a8c3c5c0e274c16b38005b4c3851f242022-12-21T22:09:18ZengFrontiers Media S.A.Frontiers in Molecular Biosciences2296-889X2021-05-01810.3389/fmolb.2021.647915647915Protein Docking Model Evaluation by Graph Neural NetworksXiao Wang0Sean T. Flannery1Daisuke Kihara2Daisuke Kihara3Department of Computer Science, Purdue University, West Lafayette, IN, United StatesDepartment of Computer Science, Purdue University, West Lafayette, IN, United StatesDepartment of Computer Science, Purdue University, West Lafayette, IN, United StatesDepartment of Biological Sciences, Purdue University, West Lafayette, IN, United StatesPhysical interactions of proteins play key functional roles in many important cellular processes. To understand molecular mechanisms of such functions, it is crucial to determine the structure of protein complexes. To complement experimental approaches, which usually take a considerable amount of time and resources, various computational methods have been developed for predicting the structures of protein complexes. In computational modeling, one of the challenges is to identify near-native structures from a large pool of generated models. Here, we developed a deep learning–based approach named Graph Neural Network–based DOcking decoy eValuation scorE (GNN-DOVE). To evaluate a protein docking model, GNN-DOVE extracts the interface area and represents it as a graph. The chemical properties of atoms and the inter-atom distances are used as features of nodes and edges in the graph, respectively. GNN-DOVE was trained, validated, and tested on docking models in the Dockground database and further tested on a combined dataset of Dockground and ZDOCK benchmark as well as a CAPRI scoring dataset. GNN-DOVE performed better than existing methods, including DOVE, which is our previous development that uses a convolutional neural network on voxelized structure models.https://www.frontiersin.org/articles/10.3389/fmolb.2021.647915/fullprotein dockingdocking model evaluationgraph neural networksdeep learningprotein structure prediction |
| spellingShingle | Xiao Wang Sean T. Flannery Daisuke Kihara Daisuke Kihara Protein Docking Model Evaluation by Graph Neural Networks Frontiers in Molecular Biosciences protein docking docking model evaluation graph neural networks deep learning protein structure prediction |
| title | Protein Docking Model Evaluation by Graph Neural Networks |
| title_full | Protein Docking Model Evaluation by Graph Neural Networks |
| title_fullStr | Protein Docking Model Evaluation by Graph Neural Networks |
| title_full_unstemmed | Protein Docking Model Evaluation by Graph Neural Networks |
| title_short | Protein Docking Model Evaluation by Graph Neural Networks |
| title_sort | protein docking model evaluation by graph neural networks |
| topic | protein docking docking model evaluation graph neural networks deep learning protein structure prediction |
| url | https://www.frontiersin.org/articles/10.3389/fmolb.2021.647915/full |
| work_keys_str_mv | AT xiaowang proteindockingmodelevaluationbygraphneuralnetworks AT seantflannery proteindockingmodelevaluationbygraphneuralnetworks AT daisukekihara proteindockingmodelevaluationbygraphneuralnetworks AT daisukekihara proteindockingmodelevaluationbygraphneuralnetworks |