Link to bioRxiv paper: http://biorxiv.org/cgi/content/short/2020.09.11.293043v1?rss=1 Authors: Orioli, S., Henning Hansen, C. G., Arleth, L. Abstract: We introduce a new software, called Marbles, that employs SAXS intensities to predict the shape of membrane proteins embedded into membrane nanodiscs. To gain computational speed and efficient convergence, the strategy is based on a hybrid approach that allows one to account for the nanodisc contribution to the SAXS intensity through a semi-analytical model, while the embedded membrane protein is treated as set of beads, similarly to well known ab-initio methods. The code, implemented in C++ with a Python user interface, provides a good performance and includes the possibility to systematically treat unstructured domains. We prove the reliability and flexibility of our approach by benchmarking the code on a toy model and two proteins of very different geometry and size. Copy rights belong to original authors. Visit the link for more info