You are here

Iterative local rebuilding vs automated structure refinement

5 posts / 0 new
Last post
Iterative local rebuilding vs automated structure refinement

I've been using the iterative local rebuilding approach based on this tutorial:

In a recent publication titled "Automated structure refinement of macromolecular assemblies from cryo-EM maps using Rosetta", the following was mentioned:

While previously we have developed an iterative local rebuilding tool that is capable of refining homology models into near-atomic-resolution cryo-EM maps (DiMaio et al., 2015), several advances were required to extend this tool so that it can successfully refine hand-built models. Our new approach includes a method for automatically detecting and correcting problematic residues in hand-built models without overfitting, a model-selection method for identifying models with good agreement to the density data and with physically realistic geometry, a voxel-size refinement method for correcting errors in calibrating the magnification scaling factor of a microscope, an improved sidechain-optimization method to correct sidechain placement errors in very large systems, and a way to estimate uncertainty in a refined model.


I wonder though how I can use the new method, because I couldn't find the XML scripts for the described protocol in the supplementary as quoted below. Also every .gz archive in the supplementary contain only 1 file with no extension, I'm not sure how to use that! Lastly, the link to the pdf below seems to be down.  

All methods described are available as part of Rosetta Software Suite, using weekly releases after week 35, 2016. The Rosetta XML files and flags for running all the refinements discussed in this manuscript are included as Supplementary files 1–3. The scripts and the tutorial used for running the method described here is available at the website of the corresponding author (


I would appreciate your input. 

Post Situation: 
Fri, 2018-09-14 07:44

I would appreciate your response.

Thu, 2018-09-20 08:28

It looks like the supplemental files at  -- despite being named *.gz -- are actually tar-gzipped files. You can rename them to *.tar.gz or *.tgz and then use standard Unix archive tools to open them.

Regarding the PDF, if you do some "URL hacking" and go up to the containing directory there's a number of additional files there, including what appears to be some updated tutorials.

Thu, 2018-09-20 08:55

Sorry for the late reply, but I still can't find the xml file. 

Tue, 2018-11-06 08:57

Which XML are you looking for? If I click on "Download elife-17219-supp1-v2.gz" at the bottom of , then rename the downloaded file to elife-17219-supp1-v2.tar.gz  and then extract that with the tar application, I can find two different refine.xml in the Wang-DiMaio-2016-Suppl-Data-1/Step1/ directory and the Wang-DiMaio-2016-Suppl-Data-1/Step2/ directory.

I'll note that the format of those XMLs are out of date with the current version of Rosetta. You'll either need to use the Rosetta/tools/xsd_xrw/ script, or better yet, look to the updated version in (e.g. in ) note that you'll probably want to read the updated tutorial PDFs in that directory, as the format/layout/method has been updated from the paper, so there isn't necessarily a 1-to-1 correspondence between the XML scripts in the paper and the ones in the new tutorial.

Wed, 2018-11-21 14:11