Re: Responding to the OMX oil comments.....

Posted by George McNamara on
URL: http://confocal-microscopy-list.275.s1.nabble.com/what-fluorescent-proteins-are-working-best-for-precision-localization-microscopy-nanoscopy-in-your-a-tp7580157p7580181.html

*****
To join, leave or search the confocal microscopy listserv, go to:
http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
*****

Hi Alison,

Thanks for clarifying for me about refractive index per specimen (I did
mention that even across a specimen the mountant R.I. can change) - and
for pointing out the room temperature and critical fluorescent channel
are both critical issues as well.

SoftWoRx - as far as I can see, Applied Precision / General Electric,
has failed to "nail the landing" on the
acquisition/recontruction/registration. Every step that I have to do
manually, they could have automated years ago. probably the only way
they will go to automation is when they start to lose sales because
someone has a much more efficient workflow.

I do agree with you that critical evaluation of the data is even more
important for structured illumination (and the other nanoscopes) than it
has been for confocal, deconvolution, and even plain old widefield.

Thanks for the tip that I also need to learn to look at - and understand
- the log files.

Sincerely,

George


On 4/12/2013 9:51 AM, Alison North wrote:

> *****
> To join, leave or search the confocal microscopy listserv, go to:
> http://lists.umn.edu/cgi-bin/wa?A0=confocalmicroscopy
> *****
>
> Dear George,
>
> You are absolutely correct that the refractive index of the immersion
> oil is critical for 3D-SIM, which is why I feel I have to respond to
> your posting and expand it a bit, in case any new or potential OMX
> users out there get confused.  You say that the 1.512 oil was "the
> best oil" - I think it needs to be clarified that you mean this was
> the best oil for that particular sample, looking at the particular
> wavelength fluorochromes you were imaging, and working at the exact
> temperature of that room.   In picking the oil for any OMX experiment
> you need to first check the room temp (if only ours were perfectly
> stable that wouldn't be an issue, but sadly we do get 1 degree
> fluctuations from day to day depending on the weather etc., despite
> having stipulated that our engineers needed to give me the most stable
> room environment possible!).  Then you need to consider the sample
> prep - by encouraging people to use the high performance coverslips
> and to try to stick to ProLong Gold (talking about fixed samples only
> of course) we aim to minimize the amount of time we need to optimize
> the oil r.i. each time.  But you will still need to decide which is
> the most critical channel in your experiment - if you are most
> interested in the red signal then you will pick a different oil than
> if the blue or green signals are the most critical, as you won't get a
> perfectly optimal signal in every channel with just one oil.
>
> I also think you are being a tad unfair on SoftWorx - I can't say I
> find it at all painful to use.  Every software has its glitches and
> annoying features but of the many different softwares we have in our
> facility we actually find SoftWorx to be one of the most user-friendly
> - and it hardly ever crashes, which is a big plus in my book.  I don't
> know which tedious manual steps you are referring to, but the more I
> have worked with the system, the more I want to slow the users down
> and force them to look at each log file while the reconstruction is
> going along.  Checking numbers in the logs like the line spacing, the
> amplitude, and the ko angles gives a good initial indication of
> whether you have a decent data set or not.  If our users just ran
> everything as automatic batch files I fear that the number of
> artifacts published would shoot up.
>
> Anyway, that's just my two cents and I expect you will disagree
> vehemently! :-)
> All the best,
> Alison
>
>
>>
>> George
>> p.s. I had training on M.D. Anderson Cancer Center's OMX yesterday. A
>> couple of observations (which may not be new to the listserv):
>> * refractive index of the immersion oil matters a lot. We (Tomasz and
>> Anna Zal were the trainers) had best results with one of Michael
>> Davidson's triple immunofluorescence slides mounted in CytoSeal.
>> Olympus 100x/1.4 NA lens. The best oil was RI=1.512. The RI=1.510 and
>> 1.514 oils were noticably worse in the reconstructions. The RI=1.518
>> was obviously a wrong choice from the acquisition screen (no or low
>> contrast fringes).
>> * Michael's DAPI in CytoSeal photoconverted to green fluorescence
>> after prolonged excitation with 405 nm laser (several of us have
>> seen, and mentioned on listserv, this with Hoechst ... I suppose it
>> is possible the slide was mislabeled and was actually Hoechst???).
>> * Prolong Gold (on Zeiss 18x18 mm coverglass) ... looks like the R.I.
>> changes from the edge to the middle of the coverglass. This makes
>> sense in that the stuff solidifies by outgasssing a [relatively low
>> refractive index] solvent. Since 0.002 steps in oil matter, it
>> appears that the Prlong Gold R.I. changes by at least 0.002 (maybe a
>> lot more) from edge to middle. This suggests that instead of spending
>> time to change the oil for different areas (distances), that it may
>> be more efficient to select an oil and then search the coverglass for
>> optimum R.I. match (best fring contrast and then final check on the
>> reconstructed image). ... I may decide to use Prolong Gold with "open
>> face" imaging dishes, or buy and test CytoSeal etc.
>> * The Applied Precision reconstruction software is called "SoftWoRx"
>> but a much more accurate name is "painfully tedious manual steps to
>> do everything anti-productivity WoRx".
>> * Plan V: I am scheduled for training on M.D. Anderson's Vutara
>> SR-200 on Monday (Tomasz and Anna doing the training again). Maybe
>> I'll end up optimizing sample preps for this instead of the OMX.
>>
>>
>