One more thing:
I got some warnings during refinement:
Are your groups large enough? Or is the reference on the correct greyscale?
3.42/4.17 min .................................................~~(,_,">
WARNING: norm_correction= 11.3176 for particle 4845 in group 3; Are your
groups large enough? Or is the reference on the correct greyscale?
3.48/4.17 min ..................................................~~(,_,">
WARNING: norm_correction= 11.3181 for particle 4820 in group 3; Are your
groups large enough? Or is the reference on the correct greyscale?
3.63/4.18 min
....................................................~~(,_,"> WARNING:
norm_correction= 14.6129 for particle 4805 in group 3; Are your groups
large enough? Or is the reference on the correct greyscale?
can that be the reason for the distorted FSC curves? I now regrouped the
particle images and, at present, refinement is running without warnings!
I'll report the final outcome!
best, Dieter
------------------------------------------------------------------------
Dieter Blaas,
Max Perutz Laboratories
Medical University of Vienna,
Inst. Med. Biochem., Vienna Biocenter (VBC),
Dr. Bohr Gasse 9/3,
A-1030 Vienna, Austria,
Tel: 0043 1 4277 61630,
Mobile: 0043 699 1942 1659
e-mail: [log in to unmask]
------------------------------------------------------------------------
On 16.02.2021 01:16, Basil Greber wrote:
> Do your masks have any sharp edges? Are they large/soft enough? Are they zero all the way to the corners of the box, without artefacts near the box edges/corners?
>
> Basil
>
>
>> Am 16.02.2021 um 00:02 schrieb Dieter Blaas <[log in to unmask]>:
>>
>> Hi,
>>
>> postprocess.star shows a reasonable resolution:
>>
>> _rlnFinalResolution 3.386336
>>
>> but the FSC curves are indeed screwed and show a bump. The curves start normally to go down but then go up again at about 1/A = 0.4. This reminds me a lot of the error I saw about a year ago....
>>
>> I am quite sure to not have duplicates....
>>
>> best, Dieter
>>
>> ------------------------------------------------------------------------
>> Dieter Blaas,
>> Max Perutz Laboratories
>> Medical University of Vienna,
>> Inst. Med. Biochem., Vienna Biocenter (VBC),
>> Dr. Bohr Gasse 9/3,
>> A-1030 Vienna, Austria,
>> Tel: 0043 1 4277 61630,
>> Mobile: 0043 699 1942 1659
>> e-mail: [log in to unmask]
>> ------------------------------------------------------------------------
>>
>> On 15.02.2021 22:12, Takanori Nakane wrote:
>>> Hi,
>>>
>>> This is strange. You should trust the value from PostProcess.
>>>
>>> Refine3D reports the HIGHEST resolution where the FSC
>>> crosses 0.143 (or 0.5 during iterations), while PostProcess
>>> reports the LOWEST resolution. If your FSC curve has
>>> oscillations or bumps, Refine3D can give too high resolution.
>>>
>>> Don't you have oscillations in the FSC?
>>> Did you make sure you don't have duplicated particles?
>>>
>>> Best regards,
>>>
>>> Takanori Nakane
>>>
>>> On 2021/02/15 19:00, Dieter Blaas wrote:
>>>> Sorry to come back to this issue but is it OK that this value is also output in run.out?
>>>>
>>>> .........................
>>>>
>>>> Calculating solvent-corrected gold-standard FSC ...
>>>> + randomize phases beyond: 4.34912 Angstroms
>>>> Maximization ...
>>>> 000/??? sec ~~(,_,"> [oo]^M3.08/3.08 mi$
>>>> Auto-refine: Refinement has converged, stopping now...
>>>> Auto-refine: + Final reconstruction from all particles is saved as: Refine3D/job030/run_cla$
>>>> Auto-refine: + Final model parameters are stored in: Refine3D/job030/run_model.star
>>>> *Auto-refine: + Final data parameters are stored in: Refine3D/job030/run_data.star**
>>>> ** Auto-refine: + Final resolution (already with masking) is: 1.9804**
>>>> *---------------------------
>>>>
>>>> Best, Dieter
>>>>
>>>> ------------------------------------------------------------------------
>>>> Dieter Blaas,
>>>> Max Perutz Laboratories
>>>> Medical University of Vienna,
>>>> Inst. Med. Biochem., Vienna Biocenter (VBC),
>>>> Dr. Bohr Gasse 9/3,
>>>> A-1030 Vienna, Austria,
>>>> Tel: 0043 1 4277 61630,
>>>> Mobile: 0043 699 1942 1659
>>>> e-mail:[log in to unmask]
>>>> ------------------------------------------------------------------------
>>>>
>>>> On 15.02.2021 16:57, Takanori Nakane wrote:
>>>>> Hi,
>>>>>
>>>>> You are looking at wrong items.
>>>>>
>>>>> rlnCurrentResolution and rlnCurrentImageSize in model.star's
>>>>> "model_general" table are those used for alignment and backprojection.
>>>>> Since "the last iteration will use data to Nyquist frequency",
>>>>> the value goes up.
>>>>>
>>>>> You should look at rlnEstimatedResolution in "model_classes" table.
>>>>>
>>>>> Best regards,
>>>>>
>>>>> Takanori Nakane
>>>>>
>>>>> On 2021/02/15 15:50, Dieter Blaas wrote:
>>>>>> Hi,
>>>>>>
>>>>>> some time ago people reported that Postprocess reported erroneous resolution values and distorted FSC curves due to an incorrect b-factor determined for single particles (I hope I am right here). Is this still the case in the above version of relion? I used CtfRefine followed by Polishing and, on re-running the refinement with shiny.dat, I get the below output on "cat *mod*|grep Curr" reporting a erroneous resolution for the last step (i.e.combining the two half maps). Is this error related to the above one? Why is it occurring? Did I use wrong parameters? The map looks OK but does certainly not have 2 A resolution!
>>>>>>
>>>>>> I did not see this problem in another refinement following the same procedure....
>>>>>>
>>>>>> .................................
>>>>>>
>>>>>> _rlnCurrentResolution 3.759407
>>>>>> _rlnCurrentImageSize 348
>>>>>> _rlnCurrentResolution 3.759407
>>>>>> _rlnCurrentImageSize 348
>>>>>> _rlnCurrentResolution 1.980402
>>>>>> _rlnCurrentImageSize 450
>>>>>>
>>>>>> Thanks, best, Dieter
>>>>>>
>>>>>> ------------------------------------------------------------------------
>>>>>> Dieter Blaas,
>>>>>> Max Perutz Laboratories
>>>>>> Medical University of Vienna,
>>>>>> Inst. Med. Biochem., Vienna Biocenter (VBC),
>>>>>> Dr. Bohr Gasse 9/3,
>>>>>> A-1030 Vienna, Austria,
>>>>>> Tel: 0043 1 4277 61630,
>>>>>> Mobile: 0043 699 1942 1659
>>>>>> e-mail: [log in to unmask]
>>>>>> ------------------------------------------------------------------------
>>>>>>
>>>>>> ########################################################################
>>>>>>
>>>>>> To unsubscribe from the CCPEM list, click the following link:
>>>>>> https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCPEM&A=1
>>>>>>
>>>>>> This message was issued to members of www.jiscmail.ac.uk/CCPEM, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
>>>> ------------------------------------------------------------------------------------------------------------------------------------------------------
>>>>
>>>> To unsubscribe from the CCPEM list, click the following link:
>>>> https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCPEM&A=1 <https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCPEM&A=1>
>>>>
>> ########################################################################
>>
>> To unsubscribe from the CCPEM list, click the following link:
>> https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCPEM&A=1
>>
>> This message was issued to members of www.jiscmail.ac.uk/CCPEM, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
########################################################################
To unsubscribe from the CCPEM list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/WA-JISC.exe?SUBED1=CCPEM&A=1
This message was issued to members of www.jiscmail.ac.uk/CCPEM, a mailing list hosted by www.jiscmail.ac.uk, terms & conditions are available at https://www.jiscmail.ac.uk/policyandsecurity/
|