Direct Inference of Base-Pairing Probabilities with Neural Networks Improves Prediction of RNA Secondary Structures with Pseudoknots
Existing approaches to predicting RNA secondary structures depend on how the secondary structure is decomposed into substructures, that is, the <i>architecture</i>, to define their parameter space. However, architecture dependency has not been sufficiently investigated, especially for ps...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2022-11-01
|
| Series: | Genes |
| Subjects: | |
| Online Access: | https://www.mdpi.com/2073-4425/13/11/2155 |
| _version_ | 1797465216318963712 |
|---|---|
| author | Manato Akiyama Yasubumi Sakakibara Kengo Sato |
| author_facet | Manato Akiyama Yasubumi Sakakibara Kengo Sato |
| author_sort | Manato Akiyama |
| collection | DOAJ |
| description | Existing approaches to predicting RNA secondary structures depend on how the secondary structure is decomposed into substructures, that is, the <i>architecture</i>, to define their parameter space. However, architecture dependency has not been sufficiently investigated, especially for pseudoknotted secondary structures. In this study, we propose a novel algorithm for directly inferring base-pairing probabilities with neural networks that do not depend on the architecture of RNA secondary structures, and then implement this approach using two maximum expected accuracy (MEA)-based decoding algorithms: Nussinov-style decoding for pseudoknot-free structures and IPknot-style decoding for pseudoknotted structures. To train the neural networks connected to each base pair, we adopt a max-margin framework, called structured support vector machines (SSVM), as the output layer. Our benchmarks for predicting RNA secondary structures with and without pseudoknots show that our algorithm outperforms existing methods in prediction accuracy. |
| first_indexed | 2024-03-09T18:18:20Z |
| format | Article |
| id | doaj.art-1f716edef3c1404db301f476e02f7cf5 |
| institution | Directory Open Access Journal |
| issn | 2073-4425 |
| language | English |
| last_indexed | 2024-03-09T18:18:20Z |
| publishDate | 2022-11-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Genes |
| spelling | doaj.art-1f716edef3c1404db301f476e02f7cf52023-11-24T08:27:09ZengMDPI AGGenes2073-44252022-11-011311215510.3390/genes13112155Direct Inference of Base-Pairing Probabilities with Neural Networks Improves Prediction of RNA Secondary Structures with PseudoknotsManato Akiyama0Yasubumi Sakakibara1Kengo Sato2Department of Biosciences and Informatics, Keio University, 3-14-1 Hiyoshi, Kohoku-ku, Yokohama 223-8522, JapanDepartment of Biosciences and Informatics, Keio University, 3-14-1 Hiyoshi, Kohoku-ku, Yokohama 223-8522, JapanSchool of System Design and Technology, Tokyo Denki University, 5 Senju Asahi-cho, Adachi-ku, Tokyo 120-8551, JapanExisting approaches to predicting RNA secondary structures depend on how the secondary structure is decomposed into substructures, that is, the <i>architecture</i>, to define their parameter space. However, architecture dependency has not been sufficiently investigated, especially for pseudoknotted secondary structures. In this study, we propose a novel algorithm for directly inferring base-pairing probabilities with neural networks that do not depend on the architecture of RNA secondary structures, and then implement this approach using two maximum expected accuracy (MEA)-based decoding algorithms: Nussinov-style decoding for pseudoknot-free structures and IPknot-style decoding for pseudoknotted structures. To train the neural networks connected to each base pair, we adopt a max-margin framework, called structured support vector machines (SSVM), as the output layer. Our benchmarks for predicting RNA secondary structures with and without pseudoknots show that our algorithm outperforms existing methods in prediction accuracy.https://www.mdpi.com/2073-4425/13/11/2155RNA secondary structuredeep learningpseudoknots |
| spellingShingle | Manato Akiyama Yasubumi Sakakibara Kengo Sato Direct Inference of Base-Pairing Probabilities with Neural Networks Improves Prediction of RNA Secondary Structures with Pseudoknots Genes RNA secondary structure deep learning pseudoknots |
| title | Direct Inference of Base-Pairing Probabilities with Neural Networks Improves Prediction of RNA Secondary Structures with Pseudoknots |
| title_full | Direct Inference of Base-Pairing Probabilities with Neural Networks Improves Prediction of RNA Secondary Structures with Pseudoknots |
| title_fullStr | Direct Inference of Base-Pairing Probabilities with Neural Networks Improves Prediction of RNA Secondary Structures with Pseudoknots |
| title_full_unstemmed | Direct Inference of Base-Pairing Probabilities with Neural Networks Improves Prediction of RNA Secondary Structures with Pseudoknots |
| title_short | Direct Inference of Base-Pairing Probabilities with Neural Networks Improves Prediction of RNA Secondary Structures with Pseudoknots |
| title_sort | direct inference of base pairing probabilities with neural networks improves prediction of rna secondary structures with pseudoknots |
| topic | RNA secondary structure deep learning pseudoknots |
| url | https://www.mdpi.com/2073-4425/13/11/2155 |
| work_keys_str_mv | AT manatoakiyama directinferenceofbasepairingprobabilitieswithneuralnetworksimprovespredictionofrnasecondarystructureswithpseudoknots AT yasubumisakakibara directinferenceofbasepairingprobabilitieswithneuralnetworksimprovespredictionofrnasecondarystructureswithpseudoknots AT kengosato directinferenceofbasepairingprobabilitieswithneuralnetworksimprovespredictionofrnasecondarystructureswithpseudoknots |