Prospects for recurrent neural network models to learn RNA biophysics from high-throughput data

Abstract

RNA is a functionally versatile molecule that plays key roles in genetic regulation and in emerging technologies to control biological processes. Computational models of RNA secondary structure are well-developed but often fall short in making quantitative predictions of the behavior of multi-RNA complexes. Recently, large datasets characterizing hundreds of thousands of individual RNA complexes have emerged as rich sources of information about RNA energetics. Meanwhile, advances in machine learning have enabled the training of complex neural networks from large datasets. Here, we assess whether a recurrent neural network model, Ribonet, can learn from high-throughput binding data, using simulation and experimental studies to test model accuracy but also determine if they learned meaningful information about the biophysics of RNA folding. We began by evaluating the model on energetic values predicted by the Turner model to assess whether the neural network could learn a representation that recovered known biophysical principles. First, we trained Ribonet to predict the simulated free energy of an RNA in complex with multiple input RNAs. Our model accurately predicts free energies of new sequences but also shows evidence of having learned base pairing information, as assessed by in silico double mutant analysis. Next, we extended this model to predict the simulated affinity between an arbitrary RNA sequence and a reporter RNA. While these more indirect measurements precluded the learning of basic principles of RNA biophysics, the resulting model achieved sub-kcal/mol accuracy and enabled design of simple RNA input responsive riboswitches with high activation ratios predicted by the Turner model from which the training data were generated. Finally, we compiled and trained on an experimental dataset comprising over 600,000 experimental affinity measurements published on the Eterna open laboratory. Though our tests revealed that the model likely did not learn a physically realistic representation of RNA interactions, it nevertheless achieved good . CC-BY-NC-ND 4.0 International license peer-reviewed) is the author/funder. It is made available under aThe copyright holder for this preprint (which was not . http://dx.doi.org/10.1101/227611 doi: bioRxiv preprint first posted online Dec. 1, 2017; 2 performance of 0.76 kcal/mol on test sets with the application of transfer learning and novel sequence-specific data augmentation strategies. These results suggest that recurrent neural network architectures, despite being naïve to the physics of RNA folding, have the potential to capture complex biophysical information. However, more diverse datasets, ideally involving more direct free energy measurements, may be necessary to train de novo predictive models that are consistent with the fundamentals of RNA biophysics. Author SummaryThe precise design of RNA interactions is essential to gaining greater control over RNA-based biotechnology tools, including designer riboswitches and CRISPR-Cas9 gene editing. However, the ...

Compact RNA sensors for increasingly complex functions of multiple inputs

“…Data compiled at the finer level of individual RNA-MaP sequence clusters are available at https://github.com/eternagame/EternaDataRibonet and described in reference. 64…”

RNA Engineering for Public Health: Innovations in RNA-Based Diagnostics and Therapeutics

“…Improving our ability to computationally design structures speeds the design-build-test cycle by providing more accurate predictions of RNA function. Crowdsourcing of structure prediction offers a method to overcome the limitations of existing computational tools to design more complex or dynamic structures (147)(148)(149)(150)(151). Furthermore, machine-learning tools can facilitate the high-throughput design of diagnostics against arbitrary sequence targets (152).…”