Parallelization of reconstruction algorithms in three-dimensional electron microscopy

Structure determination of biological specimens is one of the key issues addressed in biosciences because the functionality of biological machinery is strongly linked to its spatial structure. Although microscopes present limitations in terms of scales, indirect approaches for electron microscopy have been developed to allow visualization of biological macromolecules at nearly atomic resolution. Powerful mathematical algorithms allow researchers to obtain three-dimensional models of biological specimens. Nevertheless, those algorithms imply huge computational costs, which prevent them to be broadly used to study the structure of biological specimens. In this work we compare some of those algorithms and propose a parallelization strategy to help overcome their demands of computational power.