A Caltech Library Service

Machine learning-guided channelrhodopsin engineering enables minimally-invasive optogenetics

Bedbrook, Claire N. and Yang, Kevin K. and Robinson, J. Elliott and Mackey, Elisha D. and Gradinaru, Viviana and Arnold, Frances H. (2019) Machine learning-guided channelrhodopsin engineering enables minimally-invasive optogenetics. Nature Methods, 16 (11). pp. 1176-1184. ISSN 1548-7091. PMCID PMC6858556.

[img] PDF - Accepted Version
See Usage Policy.

[img] PDF - Submitted Version
See Usage Policy.

[img] PDF (Supplementary Figs. 1–11 and Tables 1–5) - Supplemental Material
See Usage Policy.

[img] PDF (Reporting Summary) - Supplemental Material
See Usage Policy.

[img] MS Excel (Supplementary Table 6) - Supplemental Material
See Usage Policy.

[img] MS Excel (Supplementary Data 1) - Supplemental Material
See Usage Policy.

[img] MS Excel (Supplementary Data 2) - Supplemental Material
See Usage Policy.

[img] MS Excel (Supplementary Data 3) - Supplemental Material
See Usage Policy.

[img] MS Excel (Supplementary Data 4) - Supplemental Material
See Usage Policy.

[img] Video (MPEG) (Supplementary Video 1) - Supplemental Material
See Usage Policy.

[img] Video (MPEG) (Supplementary Video 2) - Supplemental Material
See Usage Policy.

[img] MS Excel (Source Data Fig. 3) - Supplemental Material
See Usage Policy.

[img] MS Excel (Source Data Fig. 4) - Supplemental Material
See Usage Policy.


Use this Persistent URL to link to this item:


We engineered light-gated channelrhodopsins (ChRs) whose current strength and light sensitivity enable minimally invasive neuronal circuit interrogation. Current ChR tools applied to the mammalian brain require intracranial surgery for transgene delivery and implantation of fiber-optic cables to produce light-dependent activation of a small volume of tissue. To facilitate expansive optogenetics without the need for invasive implants, our engineering approach leverages the substantial literature of ChR variants to train statistical models for the design of high-performance ChRs. With Gaussian process models trained on a limited experimental set of 102 functionally characterized ChRs, we designed high-photocurrent ChRs with high light sensitivity. Three of these, ChRger1–3, enable optogenetic activation of the nervous system via systemic transgene delivery. ChRger2 enables light-induced neuronal excitation without fiber-optic implantation; that is, this opsin enables transcranial optogenetics.

Item Type:Article
Related URLs:
URLURL TypeDescription ReadCube access CentralArticle Paper
Bedbrook, Claire N.0000-0003-3973-598X
Yang, Kevin K.0000-0001-9045-6826
Robinson, J. Elliott0000-0001-9417-3938
Gradinaru, Viviana0000-0001-5868-348X
Arnold, Frances H.0000-0002-4027-364X
Additional Information:© 2019 Springer Nature Limited. Received 25 September 2018; Accepted 22 August 2019; Published 14 October 2019. Data availability: The authors declare that data supporting the findings of this study are available within the paper and its Supplementary information files. Source data for classification model training are provided in Supplementary Data 1 and 2. Source data for regression model training are provided in Supplementary Data 2. DNA constructs for the ChRger variants are deposited for distribution at Addgene (, plasmid numbers 127237-44). Code availability: Code used to train classification and regression models can be found at We thank Twist Bioscience for synthesizing and cloning ChR sequences, D. Wagenaar (California Institute of Technology) and the Caltech Neurotechnology Center for building the mouse treadmill, J. Brake (California Institute of Technology) for performing spectrometer measurements, J. Bedbrook for critical reading of the manuscript and the Gradinaru and Arnold laboratories for helpful discussions. This work was funded by the Institute for Collaborative Biotechnologies grant no. W911NF-09-0001 from the US Army Research Office (F.H.A) and the National Institutes of Health (NIH) (V.G.): NIH BRAIN grant no. RF1MH117069, NIH Director’s Pioneer Award grant no. DP1NS111369, NIH Director’s New Innovator Award grant no. DP2NS087949 and SPARC grant no. OT2OD023848. Additional funding includes the NSF NeuroNex Technology Hub grant no. 1707316 (V.G.), the CZI Neurodegeneration Challenge Network (V.G.), the Vallee Foundation (V.G.), the Heritage Medical Research Institute (V.G.) and the Beckman Institute for CLARITY, Optogenetics and Vector Engineering Research for technology development and broad dissemination: (V.G.). The content of the information does not necessarily reflect the position or the policy of the Government, and no official endorsement should be inferred. C.N.B. is funded by Ruth L. Kirschstein National Research Service Awards grant no. F31MH102913. J.E.R. is supported by the Children’s Tumor Foundation (Young Investigator Award grant no. 2016-01-006). Author Contributions: C.N.B., K.K.Y., V.G. and F.H.A. conceptualized the project. C.N.B. coordinated all experiments and data analysis. C.N.B. and K.K.Y. built machine-learning models. C.N.B. performed construct design and cloning. C.N.B. and E.D.M. performed AAV production. E.D.M. prepared cultured neurons. C.N.B. and J.E.R. conducted electrophysiology. C.N.B. and J.E.R. performed injections. J.E.R. performed fiber cannula implants and behavioral experiments. C.N.B. performed all data analysis. C.N.B. wrote the manuscript with input and editing from all authors. V.G. supervised optogenetics/electrophysiology, and F.H.A. supervised the protein engineering. Competing interests: A provisional patent application (CIT File No.: CIT-8092-P) has been filed by Caltech based on these results. C.N.B., K.K.Y., V.G. and F.H.A. are inventors on this provisional patent.
Group:Heritage Medical Research Institute
Funding AgencyGrant Number
Army Research Office (ARO)W911NF-09-0001
CZI Neurodegeneration Challenge NetworkUNSPECIFIED
Vallee FoundationUNSPECIFIED
Heritage Medical Research InstituteUNSPECIFIED
Caltech Beckman InstituteUNSPECIFIED
NIH Predoctoral FellowshipF31MH102913
Children’s Tumor Foundation2016-01-006
Subject Keywords:Machine learning; Mouse; Neuroscience; Optogenetics; Protein function predictions
Issue or Number:11
PubMed Central ID:PMC6858556
Record Number:CaltechAUTHORS:20190304-085432637
Persistent URL:
Official Citation:Bedbrook, C.N., Yang, K.K., Robinson, J.E. et al. Machine learning-guided channelrhodopsin engineering enables minimally invasive optogenetics. Nat Methods 16, 1176–1184 (2019) doi:10.1038/s41592-019-0583-8
Usage Policy:No commercial reproduction, distribution, display or performance rights in this work are provided.
ID Code:93420
Deposited By: George Porter
Deposited On:04 Mar 2019 17:16
Last Modified:30 Aug 2021 22:49

Repository Staff Only: item control page