Biocomputing Unit, Spanish National Center for Biotechnology
The cryo-electron microscopy (cryo-EM) is very quickly evolving method with growing popularity in structural biology. It allows to study specimen in its natural environment (e.g. no crystallization is needed), thus is highly relevant in many applications. The progress in electron detectors and image processing in few recent years allows cryo-EM to image specimens on nearly atomic resolution. However, the post-processing of images gathered from electron microscopes is highly computationally demanding, as the noise-to-signal ratio is very poor, the 3D specimen needs to be reconstructed from 2D images of unknown orientations and there may be several conformations of specimen observed.
A widely used tool for processing cryo-EM data is Xmipp, developed by Biocomputing Unit, Spanish National Center for Biotechnology. As the image reconstruction is highly computationally demanding, we speed-up the software to use computing facility more efficiently (activity under project World-wide E-infrastructure for structural biology - West-Life). There are several bottlenecks in image reconstruction with Xmipp, namely, image alignment and 3D Fourier reconstruction are ones of the most computationally-demanding tasks. To speedup of Xmipp, we are leveraging GPU acceleration. Currently, we have implemented a GPU-accelerated version of Fourier reconstruction, Movie alignment and image alignment in 2D/3D classification, providing an order of magnitude speedup in all cases. Currently, we are working towards optimization of workload on heterogeneous machines and utilization of autotuning. 

Results:

• David Střelák, Carlos Óscar S. Sorzano, José María Carazo, Jiří Filipovič. A GPU acceleration of 3-D Fourier reconstruction in cryo-EM. The International Journal of High Performance Computing Applications, Vol. 33, Issue 5, 2019.

This work was supported by the project EU H2020 RI West-Life - World-wide E-infrastructure for structural biology 675858.