Towards Improving Cryo-Electron Tomography Through Software Feedback

ORAL

Abstract

Cryo-electron tomography can be used to solve the structure of proteins in their natural configuration in cells. These structure solutions can form the basis of drug discovery using machine learning tools. However, the decades-old stage design necessary for tilting the thinly cut cells during tomographic data acquisition introduces large lateral and vertical shifts that make tracking difficult. The reproducibility specifications in these stage positions are a few hundreds of nanometers in scale, far from the atomic size desired in tomographic reconstructions. In moving towards better tracking, better stages need to be developed through hardware and software. A software solution is discussed and experimental results are shown indicating a path forward that may improve current stage position reproducibility by an order of magnitude.

*Data was acquired at the Electron Imaging Center for Nanosystems (EICN) at the University of California, Los Angeles's California for NanoSystems Institute (CNSI). This work was supported by the BioPACIFIC Materials Innovation Platform of the National Science Foundation under Award No. DMR-1933487.

–

Presenters

  • Nathan Joshua

    • University of California, Los Angeles

Authors

  • Nathan Joshua

    • University of California, Los Angeles

  • Khang Tran

    • University of California, Los Angeles

  • Tristan P O'Neill

    • University of California, Los Angeles (UCLA)
    • University of California, Los Angeles

  • Judy Su

    • University of California, Los Angeles