[3dem] Dose-symmetric tilt scheme for electron cryo tomography

[Ben Engel]

Hi Davide,

Counts is not electron dose, but simply a measure of the SNR of the image.
I actually don't know much more about how counts are calculated, but
perhaps David can teach us (I would be very interested to learn more!).
You must calibrate the electron dose separately, and this will be reported
in the log and recorded in the mdoc.

This #set target counts variable is the same as the variable in the normal
SerialEM tomogram acquisition.  It is an alternative to the 1/cos scheme,
where the program looks at the counts in the previous record image and then
adjusts the exposure time to try to achieve the target counts in the next
record.  If you are going up in tilt (the specimen is getting thicker),
this results in increasing exposure times.

While 1/cos works well for predictable samples, most of the people in our
lab use target counts for cellular lamellas, where it is a bit
unpredictable how quickly you will have to increase the dose, and where the
required dose increases are not the same in the plus and minus directions.

In practice, what I do is set the record for the exposure time I would like
to use in my zero tilt, take a preview image on the sample, and then take a
look in the "tilt series setup" dialogue for the predicted counts given my
exposure time.  This gives me a feeling for what I can realistically shoot
for on a given sample.  On thin cellular lamellas, I can normally get over
120 counts/image while staying under 100 e/A2 total dose.  On thin isolated
stuff, I can sometimes get up to 160 or so.  On thicker cellular lamellas,
or contaminated lamellas, I am happy to get 100 counts, and am often stuck
with only 80.  If the prediction for zero tilt is 50-60 at the maximum
electron dose I am comfortable with (1.5-2 for the image), then it simply
is too thick and I don't bother.

Best,
Ben

On Tue, Jul 11, 2017 at 5:23 PM, Davide Floris <dafloris at biophys.mpg.de>
wrote:

> Dear Ben
> after reading with a colleague the description (see forwarded message) you
> sent me, we noticed that the point 2) mentions as “CountTarget” the number
> of counts per image, while in the script it says #set target counts for
> tomogram, and the default value is set to 100.
> We are wondering:
> - is the value referred to the single image or to the whole tomogram?
> - is “counts” to be intended as “electrons” or it must be converted into
> electrons?
>
> We use a K2 summit in super resolution mode, but don’t know which camera
> was used to test the script (Falcon, CCD, K2 in linear mode?).
> Could you give us any clarification?
> Best,
>
> Davide
> _________________________________________________
>
> Davide Floris
> PhD student
> Max Planck Institute of Biophysics
> Structural Biology Department
> Max-von-Laue Straße 3
> 60438 Frankfurt am Main
> Germany
>
> (L2.090): +49-69-6303-3037 <+49%2069%2063033037>
> (B2.205): +49-69-6303-3049 <+49%2069%2063033049>
>
> On 22 May 2017, at 11:27, Ben Engel <bdengel at gmail.com> wrote:
>
> *Hi Everyone, *
>
> as promised here is a version of the Hagen tilt scheme that I modified so
> that it can be started at any starting angle.
>
> For lamellae, the pre tilt is 7 º, i.e. if you’re milling at 18º, 11º
> stage tilt should be your actual ‘0º’. I routinely use a starting angle of
> 10º,
>
> and this results in virtually the same number of counts when tilting +xº
> and -xº from your actual zero.
>
>
> *The procedure to use the script is as follows:*
>
> 0) Set up your SerialEM session as usual.
>
> 1) Copy the contents of the file to an empty macro slot and change
> parameters as needed. (Once you save your settings the script will be there
> permanently)
>
> 2) Check and adjust AngleOffset (your starting angle), CountTarget (the
> number of counts you’re asking for each image), step (the angle increment)
>
>     and tilttimes (how many times the script will do +step -> -step ->
> -2*step -> +2*step, e.g. for step = 2, tilt times = 15 you will get ±60º
> around your actual ‘zero’)
>
>     Make sure you’re not exceeding the stage limits!!!
>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.ncmir.ucsd.edu/pipermail/3dem/attachments/20170711/d4503ba4/attachment-0001.html>