Cryo-electron tomography (cryo-ET) allows researchers to image cells in their native, hydrated state at the highest resolution currently possible. However, the technique has several limitations that make analyzing the data it generates time-intensive and difficult. Hand-segmenting a single tomogram can take hours to days of human effort, but the microscope can easily generate 50 or more tomograms a day. Current deep learning segmentation programs for cryo-ET do exist but are limited to segmenting one structure at a time. Here multi-slice U-Net convolutional neural networks are trained and applied to automatically segment multiple structures simultaneously within cryo-tomograms. With proper preprocessing, these networks can be robustly inferred to many tomograms without the need for training individual networks for each tomogram. This workflow dramatically improves the speed with which cryo-electron tomograms can be analyzed by cutting segmentation time down to under 30 min in most cases. Further, segmentations can be used to improve the accuracy of filament tracing within a cellular context and to rapidly extract coordinates for subtomogram averaging.

This data set was collected on a Titan Krios G3i using TFS Tomography 5.0. It was aligned and reconstructed using IMOD, and the multi-ROI was hand-segmented by the authors using Dragonfly 2022.1.

The data comes two forms.

1) An ORS object file that contains the unprocessed tomogram, the ground truth multi-ROI, and mask used to train the network used in our JOVE protocol. This is a ready-to-use format that can be imported to Dragonfly very simply.

2) A set of .MRC files containing the unprocessed tomogram (pixel size 13.6 ang/pix), and binary masks for the groud truth input classes we used to train the data. These allow users to use the dataset outside of the Dragonfly environment.

To open the ORS object, you will need the ORS Dragonfly software suite.

MRC files are readily opened in free software such as ImageJ and IMOD

ImageJ Reference

Schneider, C. A., Rasband, W. S., & Eliceiri, K. W. (2012). NIH Image to ImageJ: 25 years of image analysis. Nature Methods, 9(7), 671–675. doi:10.1038/nmeth.2089

IMOD Reference

Kremer J.R., D.N. Mastronarde and J.R. McIntosh (1996) Computer visualization of three-dimensional image data using IMOD. J. Struct. Biol. 116:71-76.