IsoNet, a deep learning software package, was developed to enhance the resolution and interpretability of cryo-electron tomography (cryoET) data by iteratively filling in missing information and improving the signal-to-noise ratio. Applied to various cryoET datasets, IsoNet reduced resolution anisotropy and allowed for visualization of previously obscured structural details in viruses, cellular organelles, and neuronal synapses, without needing sub-tomogram averaging.

None identified

This paper presents a novel deep learning-based method for improving the resolution and interpretability of cryo-electron tomography data. The method addresses the significant challenge of anisotropic resolution and low signal-to-noise ratio, which has hindered the widespread application of cryoET. The results demonstrate the potential of this method to reveal new structural details in biological samples. However, the limitations of the method are acknowledged as well, warranting further development and investigation using more diverse and challenging datasets.