[3dem] Overcoming orientation preference issue

Craig Yoshioka yoshiokc at ohsu.edu
Thu May 18 07:46:45 PDT 2017


do it anyway

On May 17, 2017, at 11:53 PM, Dmitry Lyumkis <dlyumkis at salk.edu<mailto:dlyumkis at salk.edu>> wrote:

Alas, it is too late for us, since it is post-peer review. But seems like BioRxiv is really the way to go. Perhaps this is a relevant discussion for the upcoming GRC or NRAMM workshop.


On May 17, 2017, at 10:23 PM, Gabriel Lander <glander at gmail.com<mailto:glander at gmail.com>> wrote:

Dmitry, don't make us all wait until August!
http://biorxiv.org/

On May 17, 2017, at 10:51 AM, Dmitry Lyumkis <dlyumkis at salk.edu<mailto:dlyumkis at salk.edu>> wrote:

Dear Yang,

This is a very common problem. Severely preferred orientation not only limits your ability to reconstruct an accurate map, but can also induce overfitting, sometimes rather dramatically.

We have a manuscript coming out on this very issue detailing the problems, both with respect to overfitting, and how one might be able to generally address it by tilting the stage (as both Pawel and David had mentioned). It is currently in press, but look out for it in Nature Methods in early August. Briefly, with severely preferred orientation, I would suggest using higher tilt angles. Whenever you tilt the stage, there are several things that you have to take into account: (1) higher background contrast from inherently thicker ice; (2) increased beam induced movement from tilting; (3) limitations on estimating the CTF. The first problem is inherent to the data and cannot be solved (although it can be ameliorated by collecting using high doses). The second and third problems are essentially practical in nature and your final resolution will depend on how well you can deal with them. Definitely use gold grids and not carbon grids (unless there are better substrates now that further minimize beam-induced motion???). We have had success using MotionCor2 for movie alignment and GCTF for CTF estimation (on an “individual” particle basis), but there are certainly other ways to do things. These combinations have generally been successful for us with several particle sizes, ranging from 150 kDa to Megadalton-sized, and provided near-atomic resolution for a range of sizes, including 150 kDa. For specimens adopting mild preferred orientation, tilts up to 20-30° are probably good enough; for the pathological cases, 40-50° might be what you’re looking for. If I had to venture a guess based on what you’re saying, given the fact that you have 2 preferred orientations, I would probably suggest somewhere in the 30-40° range (but maybe try 30 first, which will have less beam-induced movement, and will be easier to reach high-resolution). Some of these details are still being worked out. Please contact me directly if you would like more information.

Many of the other suggestions that have been proposed may also work. For example, detergents should help, and many people have had success, but they require higher protein concentrations to induce the sample to go into holes. In my personal experience, either with DDM or NP40, one would have to add detergent to a concentration that is ~CMC (or slightly higher). When I worked with the HIV trimer, PMC3954647<https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3954647/>, I specifically found that one can induce more random orientations at DDM concentrations of ~2x below the CMC, whereas anything lower than that would not be enough. Since then, we worked with DDM concentrations at CMC for different samples. Still, this is going to be specimen dependent, and I have not worked with Tween20.

Good luck!

Dmitry



Dmitry Lyumkis
Salk Helmsley Fellow
The Salk Institute for Biological Studies
10010 N. Torrey Pines Rd., La Jolla, CA 92037
E: dlyumkis at salk.edu<mailto:dlyumkis at salk.edu>
T: (858)-453-4100 ext. 1155

On May 17, 2017, at 9:24 AM, Jiang,Qiu-Xing <qxjiang at UFL.EDU<mailto:qxjiang at UFL.EDU>> wrote:

Dear yang,
Besides other tricks suggested here, in our hands we noticed that our chemically oxidized carbon films have different surface property from the glow-discharged films and can change the orientational distribution by introducing a different interaction profile. A detailed procedure for chemical oxidation and cleaning is available in https://www.ncbi.nlm.nih.gov/pubmed/24457027 . It might provide a different alternative.
Best luck.
Qiu-Xing

—
From: 3dem <3dem-bounces at ncmir.ucsd.edu<mailto:3dem-bounces at ncmir.ucsd.edu>> on behalf of Yang Li <yanglixtal at gmail.com<mailto:yanglixtal at gmail.com>>
Date: Wednesday, May 17, 2017 at 8:46 AM
To: "3dem at ncmir.ucsd.edu<mailto:3dem at ncmir.ucsd.edu>" <3dem at ncmir.ucsd.edu<mailto:3dem at ncmir.ucsd.edu>>
Subject: [3dem] Overcoming orientation preference issue

Dear colleagues,

We have a protein sample that suffers from severe orientation preference, that most of the particles cluster into two distinct orientations. This way we have to collect large amounts of data in order to obtain enough effective particles, which hiders us from reaching high resolution. We have tried to make thicker ice or adding tiny amount of detergent such as Tween20, but not working very well so far. I wonder if there are any tricks we can try to overcome this orientation preference issue? Thank you in advance for suggestions!

Best,
Yang
_______________________________________________
3dem mailing list
3dem at ncmir.ucsd.edu<mailto:3dem at ncmir.ucsd.edu>
https://mail.ncmir.ucsd.edu/mailman/listinfo/3dem

_______________________________________________
3dem mailing list
3dem at ncmir.ucsd.edu<mailto:3dem at ncmir.ucsd.edu>
https://mail.ncmir.ucsd.edu/mailman/listinfo/3dem


_______________________________________________
3dem mailing list
3dem at ncmir.ucsd.edu<mailto:3dem at ncmir.ucsd.edu>
https://mail.ncmir.ucsd.edu/mailman/listinfo/3dem

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ncmir.ucsd.edu/pipermail/3dem/attachments/20170518/85d3f584/attachment-0001.html>


More information about the 3dem mailing list