Hi Dimitry

I will try to be patient in my explanation... You have deleted the 
discussion history in your response but do read our bioRxiv 2017 paper 
and trace back the FSC literature from that.
I know from our own private email exchanges that you do understand the 
basic principles.  The EMD-0144 that you are referring to has a 
pixel/voxel size of 0.81A. (I have no idea who the authors are but, 
friends or foes, that is entirely irrelevant to this discussion). The 
resolution claim in that paper is 1,62A , that is, exactly the Niquist 
frequency or twice the pixel size of 0.81A!  I have already mentioned in 
this thread that the FSC 0.143 is itself wrong while based on sloppy 
statistics (bioRxiv 2017 and trace back the literature, you may also 
want to listen to my lecture at the Wiley 2017 award ceremony which is 
available on line).  I will not repeat that argumentation here.

Here I will just give you one theoretical argument why you cannot claim 
the Nyquist frequency as the reproducible resolution. The sampling 
theorem 
(https://en.wikipedia.org/wiki/Nyquist%E2%80%93Shannon_sampling_theorem) 
requires that that signal you are digitizing is band limited, implying 
that the power of the signal in Fourier space at the Nyquist frequency 
must have dropped to zero. If the power in the signal's FT is zero in 
every measurement there is no point in correlating two independent 
measurement as needed for an FSC measurement. If, on the other hand, if 
the power is NOT zero at Nyquist, the data is under-sampled, thus 
violating the sampling theorem.  This is one reason why you will find 
the 3 pixel rule in all of our FSC related papers (for example: Orlova 
et al  J. Mol. Biol. (1997) 271, 417-437).  The safety margin we defined 
in the past is 2/3 Nyquist, stick to it (Referees take heed).

Cheers,

Marin


On 02/09/2018 14:43, Dimitry Tegunov wrote:
> Dear Marin,
>
> could you please provide us with a reference for this fundamental basis in existing literature? Surely I'm not the only one who would love to learn why the seemingly correct features of e. g. EMD-0144 (http://www.ebi.ac.uk/pdbe/entry/emdb/EMD-0144) are illegally resolved.
>
> Cheers,
> Dimitry
>
> ########################################################################
>
> To unsubscribe from the CCPEM list, click the following link:
> https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCPEM&A=1


-- 
==============================================================

     Prof Dr Ir Marin van Heel

     Laboratório Nacional de Nanotecnologia - LNNano
     CNPEM/LNNano, Campinas, Brazil

     tel:    +55-19-3518-2316
     mobile  +55-19-983455450 (current)
     mobile  +55-19-981809332
                  (041-19-981809332 TIM)
     Skype:  Marin.van.Heel
     email:  marin.vanheel(A_T)gmail.com
             marin.vanheel(A_T)lnnano.cnpem.br
     and:    mvh.office(A_T)gmail.com

--------------------------------------------------
     Emeritus Professor of Cryo-EM Data Processing
     Leiden University
     Mobile NL: +31(0)652736618 (ALWAYS ACTIVE SMS)
--------------------------------------------------
     Emeritus Professor of Structural Biology
     Imperial College London
     Faculty of Natural Sciences
     email: m.vanheel(A_T)imperial.ac.uk
--------------------------------------------------

I receive many emails per day and, although I try,
there is no guarantee that I will actually read each incoming email.

########################################################################

To unsubscribe from the CCPEM list, click the following link:
https://www.jiscmail.ac.uk/cgi-bin/webadmin?SUBED1=CCPEM&A=1