Applying machine learning to predict viral assembly for adeno-associated virus capsid libraries
Machine learning (ML) can aid in novel discoveries in the field of viral gene therapy. Specifically, big data gathered through next-generation sequencing (NGS) of complex capsid libraries is an especially prominent source of lost potential in data analysis and prediction. Furthermore, adeno-associat...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Elsevier
2021-03-01
|
| Series: | Molecular Therapy: Methods & Clinical Development |
| Subjects: | |
| Online Access: | http://www.sciencedirect.com/science/article/pii/S2329050120302448 |
| _version_ | 1829533488090447872 |
|---|---|
| author | Andrew D. Marques Michael Kummer Oleksandr Kondratov Arunava Banerjee Oleksandr Moskalenko Sergei Zolotukhin |
| author_facet | Andrew D. Marques Michael Kummer Oleksandr Kondratov Arunava Banerjee Oleksandr Moskalenko Sergei Zolotukhin |
| author_sort | Andrew D. Marques |
| collection | DOAJ |
| description | Machine learning (ML) can aid in novel discoveries in the field of viral gene therapy. Specifically, big data gathered through next-generation sequencing (NGS) of complex capsid libraries is an especially prominent source of lost potential in data analysis and prediction. Furthermore, adeno-associated virus (AAV)-based capsid libraries are becoming increasingly popular as a tool to select candidates for gene therapy vectors. These higher complexity AAV capsid libraries have previously been created and selected in vivo; however, in silico analysis using ML computer algorithms may augment smarter and more robust libraries for selection. In this study, data of AAV capsid libraries gathered before and after viral assembly are used to train ML algorithms. We found that two ML computer algorithms, artificial neural networks (ANNs), and support vector machines (SVMs), can be trained to predict whether unknown capsid variants may assemble into viable virus-like structures. Using the most accurate models constructed, hypothetical mutation patterns in library construction were simulated to suggest the importance of N495, G546, and I554 in AAV2-derived capsids. Finally, two comparative libraries were generated using ML-derived data to biologically validate these findings and demonstrate the predictive power of ML in vector design. |
| first_indexed | 2024-12-16T18:57:51Z |
| format | Article |
| id | doaj.art-b2eb5e6ad8804298a15cddd438452aff |
| institution | Directory Open Access Journal |
| issn | 2329-0501 |
| language | English |
| last_indexed | 2024-12-16T18:57:51Z |
| publishDate | 2021-03-01 |
| publisher | Elsevier |
| record_format | Article |
| series | Molecular Therapy: Methods & Clinical Development |
| spelling | doaj.art-b2eb5e6ad8804298a15cddd438452aff2022-12-21T22:20:28ZengElsevierMolecular Therapy: Methods & Clinical Development2329-05012021-03-0120276286Applying machine learning to predict viral assembly for adeno-associated virus capsid librariesAndrew D. Marques0Michael Kummer1Oleksandr Kondratov2Arunava Banerjee3Oleksandr Moskalenko4Sergei Zolotukhin5Department of Pediatrics, Division of Cellular and Molecular Therapy, University of Florida, Gainesville, FL 32608, USA; Corresponding author: Andrew D. Marques, Department of Pediatrics, Division of Cellular and Molecular Therapy, University of Florida, Gainesville, FL 32608, USA.Department of Computer & Information Science & Engineering, University of Florida, Gainesville, FL 32603, USADepartment of Pediatrics, Division of Cellular and Molecular Therapy, University of Florida, Gainesville, FL 32608, USADepartment of Computer & Information Science & Engineering, University of Florida, Gainesville, FL 32603, USAUniversity of Florida Research Computing, University of Florida, Gainesville, FL 32608, USADepartment of Pediatrics, Division of Cellular and Molecular Therapy, University of Florida, Gainesville, FL 32608, USAMachine learning (ML) can aid in novel discoveries in the field of viral gene therapy. Specifically, big data gathered through next-generation sequencing (NGS) of complex capsid libraries is an especially prominent source of lost potential in data analysis and prediction. Furthermore, adeno-associated virus (AAV)-based capsid libraries are becoming increasingly popular as a tool to select candidates for gene therapy vectors. These higher complexity AAV capsid libraries have previously been created and selected in vivo; however, in silico analysis using ML computer algorithms may augment smarter and more robust libraries for selection. In this study, data of AAV capsid libraries gathered before and after viral assembly are used to train ML algorithms. We found that two ML computer algorithms, artificial neural networks (ANNs), and support vector machines (SVMs), can be trained to predict whether unknown capsid variants may assemble into viable virus-like structures. Using the most accurate models constructed, hypothetical mutation patterns in library construction were simulated to suggest the importance of N495, G546, and I554 in AAV2-derived capsids. Finally, two comparative libraries were generated using ML-derived data to biologically validate these findings and demonstrate the predictive power of ML in vector design.http://www.sciencedirect.com/science/article/pii/S2329050120302448Machine LearningAAVCapsid LibrariesAssemblyPackagingANN |
| spellingShingle | Andrew D. Marques Michael Kummer Oleksandr Kondratov Arunava Banerjee Oleksandr Moskalenko Sergei Zolotukhin Applying machine learning to predict viral assembly for adeno-associated virus capsid libraries Molecular Therapy: Methods & Clinical Development Machine Learning AAV Capsid Libraries Assembly Packaging ANN |
| title | Applying machine learning to predict viral assembly for adeno-associated virus capsid libraries |
| title_full | Applying machine learning to predict viral assembly for adeno-associated virus capsid libraries |
| title_fullStr | Applying machine learning to predict viral assembly for adeno-associated virus capsid libraries |
| title_full_unstemmed | Applying machine learning to predict viral assembly for adeno-associated virus capsid libraries |
| title_short | Applying machine learning to predict viral assembly for adeno-associated virus capsid libraries |
| title_sort | applying machine learning to predict viral assembly for adeno associated virus capsid libraries |
| topic | Machine Learning AAV Capsid Libraries Assembly Packaging ANN |
| url | http://www.sciencedirect.com/science/article/pii/S2329050120302448 |
| work_keys_str_mv | AT andrewdmarques applyingmachinelearningtopredictviralassemblyforadenoassociatedviruscapsidlibraries AT michaelkummer applyingmachinelearningtopredictviralassemblyforadenoassociatedviruscapsidlibraries AT oleksandrkondratov applyingmachinelearningtopredictviralassemblyforadenoassociatedviruscapsidlibraries AT arunavabanerjee applyingmachinelearningtopredictviralassemblyforadenoassociatedviruscapsidlibraries AT oleksandrmoskalenko applyingmachinelearningtopredictviralassemblyforadenoassociatedviruscapsidlibraries AT sergeizolotukhin applyingmachinelearningtopredictviralassemblyforadenoassociatedviruscapsidlibraries |