beta_helix Staff Lv 1
In the spirit of fairness for the Foldit vs. UMich Electron Density Challenge, the Bardwell Lab at the University of Michigan is kindly sharing with us any additional information that they provide their students.
(A lot of this might not apply to Foldit, since the University of Michigan students are using Coot in their class, but they want to make sure to keep the competition fair by making sure everyone has the exact same information)
Here is a summary of what seemed to be working best for their students in last thursday's lab when they started working on the structure using Coot:
-
Use Find Secondary Structure (under the calculate -> other modeling tools menu) to find the most obvious alpha helices and B-strands
-
Use Ca Zone -> Mainchain (under the same menu) and click on either end of the secondary structure element to turn the ideal strands and helices into all-atom models.
-
Use Real Space Refine (under calculate -> model/fit/refine menu) to improve the fit of the secondary structure elements. Tweaking the weighting under refine/regularize control (also under model/fit/refine menu) to 10 can help to get the angles and bond parameters better.
-
Continue building the chain by iteratively using Add Terminal Residue (under model/fit/refine) and Real Space Refine on the new residues.
-
When encountering really obvious side chains (tryptophan and phenylalanine should be the easiest), model them in using Simple Mutate (also on the model/fit/refine menu), followed by rotameric adjustments (using either Rotamers or Auto Fit Rotamer on the model/fit/refine menu), followed by real space refinement. Once you have enough of these residues to get a handle on what part of the sequence you are at, use the protein sequence to start modeling in the other side chains.