Coot Speedrun Map 2 is a step up from Map 1. The goal is to create, using de novo model-building, an accurate and full atomic model for a cryo-EM reconstruction.
I fumbled the release of 0.9.4, as I was preparing the binaries I found a critical bug. So, I’ve made 0.9.4.1 a few days later. Today, that is. Here’s the combined release notes:
Here’s a bit of holiday fun: A Coot Speedrun! Pit your Coot skills against the rest of the world!
Release 0.9.3 of Coot is now available. This fixes the Colour-by-chain colours and (perhaps more importantly) fixes the problem apparent when using mmCIF files for backup/undo/redo.
This is a bug-fix release, I’d made of a mess of “colour by B-factor” colouring. This releases fixes most of that.
My colleagues recently published a paper (Nakane et al. (2020) “Single Particle cryo-EM at atomic resolution”) and asked me to make a figure. I did so and it is this week’s front cover image.
- By Ana Casañal & Paul Emsley
This tutorial is designed for 0.9.1 or later.
- By Ana Casañal & Paul Emsley
This tutorial is designed for 0.9.1 or later
Clustalw2 has now joined the Coot dependencies - that means that we can use it for alignments locally (yay!) This replaces the need to use MUSCLE on the EBI web server - which causes cut and paste kerfuffle, and that is entirely non-winning at tutorials.
Some people, maybe even myself, don’t much like having atoms in default represention. The can now be turned of globally:
I may have sent you this to this post. It’s because I am bored of typing out basically the same message every week for the last several years.
I find myself spending quite a lot of time working on the graphics system of Coot-0.9.9. There is (or has been) a lot of graphics infrasture to write/re-write/update and as I am no graphics expert, there have been many errors along the way. Some of them amused me and so I reproduce them here. Some of them are the “unselected” from a group of related images.
Sai Li and co-workers (not least of whom is Max Crispin, a recent co-author of my own) published “Molecular architecture of the SARS-CoV-2 virus” in Cell today: https://t.co/pQAbjhGh2V
Coot makes use of the CCP4 Program “Acedrg” to make links between ligands and proteins.
The Coot Utility Refinement Library Extension Wrangler (Curlew) is, the means by which one can easily install extensions.
A long-requested feature has now (eventually) been implemented. The scale for the density fit graph is now easy to find:
It turns out the the missing-residue loops feature was not found to be desirable by everyone - including bucanneer users.
Waters and ions are now drawn with a larger radius than atoms in residues.
I have head reports from Windows users that their backup file names fail to be written out. I think that this is because there are colons in the file name. To prevent this, use the following setting:
Bond orders for protein have been enabled in Colour-by-chain mode (0.9.1-pre). I am not sure that I like it yet.
Sometime Coot reports that it fails to compress the backup files in the coot-backup directory. This might be, perhaps, because the full path of the directory contains a space. In such a case you can turn off the backup compression:
At the request of John Berrisford and Oliver Clarke, I have changed the representation of loops where the inter-residue number difference is too short for the actual distance between the residue - to say it another way, there are not enough residues in the sequence to span the gap.
Hamish has recently made a demonstration video for CootVR. Slightly non-optimal video and audio, but interesting.
The data for the Ramachandran Plot in Clipper have been updated to Top8000 and these are now used in Coot. Here is an example of the atom selection in use and a plot for the Pre-PRO residue type.
Coot now displays missing loops with dotted lines when in CA Mode:
I have now enabled the reading of models with duplicate residue-numbers and insertion codes for different residues by default.
- Open a new coot, with no state script
Yay, bug fix list emptied - Here is the progress over the last couple of months:
Now we can render our model molecules with “variable” bond thickness, so that molecules that are further away have thinner bond widths (and as we zoom in, they get thicker). You might call this “Natural” Bond Width Mode.
Now you can build N-linked carbohydrate without a gui:
So now we have labels, thanks to a request from Vito Calderone.
We were looking like a release was imminent. But now, post-Madrid, there is more work to do.
Trans Proline links (PTRANS) now have planar peptide restraints applied by default.
The LINK records are now represented in the header browser:
CootVR was demonstrated to the public for the first time at the CCP4-EM Meeting at Keele University on the 11th April 2018.
subscribe via RSS