A deep learning algorithm to translate and classify cardiac electrophysiology

  1. Parya Aghasafari Ph.D.
  2. Pei-Chi Yang Ph.D.
  3. Divya C Kernik Ph.D.
  4. Kazuho Sakamoto Ph.D.
  5. Yasunari Kanda Ph.D.
  6. Junko Kurokawa Ph.D
  7. Igor Vorobyov
  8. Colleen E Clancy Ph.D.  Is a corresponding author
  1. University of California Davis, United States
  2. Washington University in St. Louis, United States
  3. University of Shizuoka, Japan
  4. National Institute of Health Sciences, Japan
  5. University California Davis, United States

Abstract

The development of induced pluripotent stem cell-derived cardiomyocytes (iPSC-CMs) has been a critical in vitro advance in the study of patient-specific physiology, pathophysiology and pharmacology. We designed a new deep learning multitask network approach intended to address the low throughput, high variability and immature phenotype of the iPSC-CM platform. The rationale for combining translation and classification tasks is because the most likely application of the deep learning technology we describe here is to translate iPSC-CMs following application of a perturbation. The deep learning network was trained using simulated action potential (AP) data and applied to classify cells into the drug-free and drugged categories and to predict the impact of electrophysiological perturbation across the continuum of aging from the immature iPSC-CMs to the adult ventricular myocytes. The phase of the AP extremely sensitive to perturbation due to a steep rise of the membrane resistance was found to contain the key information required for successful network multitasking. We also demonstrated successful translation of both experimental and simulated iPSC-CM AP data validating our network by prediction of experimental drug-induced effects on adult cardiomyocyte APs by the latter.

Data availability

Since we used simulated data, we have made all drugged and drug-free iPSC-CM and adult-CM AP data used for training and testing the multitask network publicly available at Clancy lab Github.(https://github.com/ClancyLabUCD/Multitask_network/tree/master/data). In addition, we have illustrated training and test dataset in Figure1 and Figure5.We have also shared the jupyter notebook for preparing clean and organized data for training the network at Clancy lab Github (https://github.com/ClancyLabUCD/Multitask_network/tree/master/jupyter).We also made experimental data used for the model validation publicly available at Clancy lab Github.(https://github.com/ClancyLabUCD/Multitask_network/blob/master/data/clean_data/experiments.csv ). Figure 7 illustrates the experimental data we used to validate the network.

Article and author information

Author details

  1. Parya Aghasafari Ph.D.

    Physiology and Membrane Biology, University of California Davis, Davis, United States
    Competing interests
    The authors declare that no competing interests exist.
  2. Pei-Chi Yang Ph.D.

    Physiology and Membrane Biology, University of California Davis, Davis, United States
    Competing interests
    The authors declare that no competing interests exist.
  3. Divya C Kernik Ph.D.

    Biomedical Engineering, Washington University in St. Louis, St. Louis, United States
    Competing interests
    The authors declare that no competing interests exist.
  4. Kazuho Sakamoto Ph.D.

    Bio-Informational Pharmacology, University of Shizuoka, Shizuoka, Japan
    Competing interests
    The authors declare that no competing interests exist.
  5. Yasunari Kanda Ph.D.

    Division of Pharmacology, National Institute of Health Sciences, Kanagawa, Japan
    Competing interests
    The authors declare that no competing interests exist.
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0003-2527-3526
  6. Junko Kurokawa Ph.D

    Bio-Informational Pharmacology, University of Shizuoka, Shizuoka, Japan
    Competing interests
    The authors declare that no competing interests exist.
  7. Igor Vorobyov

    University California Davis, Davis, United States
    Competing interests
    The authors declare that no competing interests exist.
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-4767-5297
  8. Colleen E Clancy Ph.D.

    Physiology and Membrane Biology, University of California Davis, Davis, United States
    For correspondence
    ceclancy@ucdavis.edu
    Competing interests
    The authors declare that no competing interests exist.
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0001-6849-4885

Funding

Common Fund (OT2OD026580)

  • Igor Vorobyov
  • Colleen E Clancy Ph.D.

Texas Advanced Computing Center Leadership Resource Allocation (MCB20010)

  • Igor Vorobyov
  • Colleen E Clancy Ph.D.

Oracle cloud for research allocation

  • Igor Vorobyov
  • Colleen E Clancy Ph.D.

Common Fund (OT2OD025308‐01S2)

  • Parya Aghasafari Ph.D.

American Heart Association (19CDA34770101)

  • Igor Vorobyov

National Heart, Lung, and Blood Institute (R01HL152681)

  • Igor Vorobyov
  • Colleen E Clancy Ph.D.

National Heart, Lung, and Blood Institute (R01HL128170)

  • Colleen E Clancy Ph.D.

National Heart, Lung, and Blood Institute (U01HL126273)

  • Colleen E Clancy Ph.D.

Department of Physiology and Membrane Biology Research Partnership Fund

  • Igor Vorobyov
  • Colleen E Clancy Ph.D.

Extreme Science and Engineering Discovery Environment (MCB170095)

  • Igor Vorobyov
  • Colleen E Clancy Ph.D.

National Center for Supercomputing Applications Blue Waters Broadening Participation Allocation

  • Igor Vorobyov
  • Colleen E Clancy Ph.D.

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Reviewing Editor

  1. Thomas Hund, The Ohio State University, United States

Version history

  1. Received: March 11, 2021
  2. Accepted: June 29, 2021
  3. Accepted Manuscript published: July 2, 2021 (version 1)
  4. Version of Record published: July 15, 2021 (version 2)

Copyright

© 2021, Aghasafari et al.

This article is distributed under the terms of the Creative Commons Attribution License permitting unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 1,845
    Page views
  • 265
    Downloads
  • 17
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. Parya Aghasafari Ph.D.
  2. Pei-Chi Yang Ph.D.
  3. Divya C Kernik Ph.D.
  4. Kazuho Sakamoto Ph.D.
  5. Yasunari Kanda Ph.D.
  6. Junko Kurokawa Ph.D
  7. Igor Vorobyov
  8. Colleen E Clancy Ph.D.
(2021)
A deep learning algorithm to translate and classify cardiac electrophysiology
eLife 10:e68335.
https://doi.org/10.7554/eLife.68335

Share this article

https://doi.org/10.7554/eLife.68335

Further reading

    1. Computational and Systems Biology
    2. Neuroscience
    Domingos Leite de Castro, Miguel Aroso ... Paulo Aguiar
    Research Article Updated

    Closed-loop neuronal stimulation has a strong therapeutic potential for neurological disorders such as Parkinson’s disease. However, at the moment, standard stimulation protocols rely on continuous open-loop stimulation and the design of adaptive controllers is an active field of research. Delayed feedback control (DFC), a popular method used to control chaotic systems, has been proposed as a closed-loop technique for desynchronisation of neuronal populations but, so far, was only tested in computational studies. We implement DFC for the first time in neuronal populations and access its efficacy in disrupting unwanted neuronal oscillations. To analyse in detail the performance of this activity control algorithm, we used specialised in vitro platforms with high spatiotemporal monitoring/stimulating capabilities. We show that the conventional DFC in fact worsens the neuronal population oscillatory behaviour, which was never reported before. Conversely, we present an improved control algorithm, adaptive DFC (aDFC), which monitors the ongoing oscillation periodicity and self-tunes accordingly. aDFC effectively disrupts collective neuronal oscillations restoring a more physiological state. Overall, these results support aDFC as a better candidate for therapeutic closed-loop brain stimulation.

    1. Cancer Biology
    2. Computational and Systems Biology
    Sara Latini, Veronica Venafra ... Francesca Sacco
    Research Article

    Currently, the identification of patient-specific therapies in cancer is mainly informed by personalized genomic analysis. In the setting of acute myeloid leukemia (AML), patient-drug treatment matching fails in a subset of patients harboring atypical internal tandem duplications (ITDs) in the tyrosine kinase domain of the FLT3 gene. To address this unmet medical need, here we develop a systems-based strategy that integrates multiparametric analysis of crucial signaling pathways, and patient-specific genomic and transcriptomic data with a prior knowledge signaling network using a Boolean-based formalism. By this approach, we derive personalized predictive models describing the signaling landscape of AML FLT3-ITD positive cell lines and patients. These models enable us to derive mechanistic insight into drug resistance mechanisms and suggest novel opportunities for combinatorial treatments. Interestingly, our analysis reveals that the JNK kinase pathway plays a crucial role in the tyrosine kinase inhibitor response of FLT3-ITD cells through cell cycle regulation. Finally, our work shows that patient-specific logic models have the potential to inform precision medicine approaches.