Dataset Information

Applying machine learning to predict viral assembly for adeno-associated virus capsid libraries.

ABSTRACT: 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.

SUBMITTER: Marques AD

PROVIDER: S-EPMC7809249 | biostudies-literature | 2021 Mar

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Applying machine learning to predict viral assembly for adeno-associated virus capsid libraries.

Marques Andrew D AD Kummer Michael M Kondratov Oleksandr O Banerjee Arunava A Moskalenko Oleksandr O Zolotukhin Sergei S

Molecular therapy. Methods & clinical development 20201203

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 creat ...[more]

PMID: 33511242

Similar Datasets

Project description:The HI loop is a prominent domain on the adeno-associated virus (AAV) capsid surface that extends from each viral protein (VP) subunit overlapping the neighboring fivefold VP. Despite the highly conserved nature of the residues at the fivefold pore, the HI loops surrounding this critical region vary significantly in amino acid sequence between the AAV serotypes. In order to understand the role of this unique capsid domain, we ablated side chain interactions between the HI loop and the underlying EF loop in the neighboring VP subunit by generating a collection of deletion, insertion, and substitution mutants. A mutant lacking the HI loop was unable to assemble particles, while a substitution mutant (10 glycine residues) assembled particles but was unable to package viral genomes. Substitution mutants carrying corresponding regions from AAV1, AAV4, AAV5, and AAV8 yielded (i) particles with titers and infectivity identical to those of AAV2 (AAV2 HI1 and HI8), (ii) particles with a decreased virus titer (1 log) but normal infectivity (HI4), and (iii) particles that synthesized VPs but were unable to assemble into intact capsids (HI5). AAV5 HI is shorter than all other HI loops by one amino acid. Replacing the missing residue (threonine) in AAV2 HI5 resulted in a moderate particle assembly rescue. In addition, we replaced the HI loop with peptides varying in length and amino acid sequence. This region tolerated seven-amino-acid peptide substitutions unless they spanned a conserved phenylalanine at amino acid position 661. Mutation of this highly conserved phenylalanine to a glycine resulted in a modest decrease in virus titer but a substantial decrease (1 log order) in infectivity. Subsequently, confocal studies revealed that AAV2 F661G is incapable of efficiently completing a key step in the infectious pathway nuclear entry, hinting at a possible perturbation of VP1 phospholipase activity. Molecular modeling studies with the F661G mutant suggest that disruption of interactions between F661 and an underlying P373 residue in the EF loop of the neighboring subunit might adversely affect incorporation of the VP1 subunit at the fivefold axis. Western blot analysis confirmed inefficient incorporation of VP1, as well as a proteolytically processed VP1 subunit that could account for the markedly reduced infectivity. In summary, our studies show that the HI loop, while flexible in amino acid sequence, is critical for AAV capsid assembly, proper VP1 subunit incorporation, and viral genome packaging, all of which implies a potential role for this unique surface domain in viral infectivity.

Project description:BackgroundIdentifying individuals who are unlikely to adhere to a physical exercise regime has potential to improve physical activity interventions. The aim of this paper is to develop and test adherence prediction models using objectively measured physical activity data in the Mobile Phone-Based Physical Activity Education program (mPED) trial. To the best of our knowledge, this is the first to apply Machine Learning methods to predict exercise relapse using accelerometer-recorded physical activity data.MethodsWe use logistic regression and support vector machine methods to design two versions of a Discontinuation Prediction Score (DiPS), which uses objectively measured past data (e.g., steps and goal achievement) to provide a numerical quantity indicating the likelihood of exercise relapse in the upcoming week. The respective prediction accuracy of these two versions of DiPS are compared, and then numerical simulation is performed to explore the potential of using DiPS to selectively allocate financial incentives to participants to encourage them to increase physical activity.Resultswe had access to a physical activity trial data that were continuously collected every 60 sec every day for 9 months in 210 participants. By using the first 15 weeks of data as training and test on weeks 16-30, we show that both versions of DiPS have a test AUC of 0.9 with high sensitivity and specificity in predicting the probability of exercise adherence. Simulation results assuming different intervention regimes suggest the potential benefit of using DiPS as a score to allocate resources in physical activity intervention programs in reducing costs over other allocation schemes.ConclusionsDiPS is capable of making accurate and robust predictions for future weeks. The most predictive features are steps and physical activity intensity. Furthermore, the use of DiPS scores can be a promising approach to determine when or if to provide just-in-time messages and step goal adjustments to improve compliance. Further studies on the use of DiPS in the design of physical activity promotion programs are warranted.Trial registrationClinicalTrials.gov NCT01280812 Registered on January 21, 2011.

Dataset Information

Applying machine learning to predict viral assembly for adeno-associated virus capsid libraries.

Publications

Applying machine learning to predict viral assembly for adeno-associated virus capsid libraries.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets