A deep learning approach to programmable RNA switches.

Journal: Nature communications
PMID: 33028812

Abstract

Engineered RNA elements are programmable tools capable of detecting small molecules, proteins, and nucleic acids. Predicting the behavior of these synthetic biology components remains a challenge, a situation that could be addressed through enhanced pattern recognition from deep learning. Here, we investigate Deep Neural Networks (DNN) to predict toehold switch function as a canonical riboswitch model in synthetic biology. To facilitate DNN training, we synthesize and characterize in vivo a dataset of 91,534 toehold switches spanning 23 viral genomes and 906 human transcription factors. DNNs trained on nucleotide sequences outperform (R = 0.43-0.70) previous state-of-the-art thermodynamic and kinetic models (R = 0.04-0.15) and allow for human-understandable attention-visualizations (VIS4Map) to identify success and failure modes. This work shows that deep learning approaches can be used for functionality predictions and insight generation in RNA synthetic biology.

Authors

  • Nicolaas M Angenent-Mari
    Department of Biological Engineering, Massachusetts Institute of Technology (MIT), Cambridge, MA, 02139, USA.
  • Alexander S Garruss
    Wyss Institute for Biologically Inspired Engineering, Harvard University, Boston, MA, 02115, USA.
  • Luis R Soenksen
    Department of Biological Engineering, Massachusetts Institute of Technology (MIT), Cambridge, MA, 02139, USA.
  • George Church
    Wyss Institute for Biologically Inspired Engineering, Harvard University, Boston, MA, 02115, USA.
  • James J Collins
    Infectious Disease and Microbiome Program, Broad Institute of MIT and Harvard, Cambridge, MA, USA.

Keywords

External Resources

Popular Topics
Recent Journals