FSC plot analysis

[Michael Kwan]

Dear all,

My FSC plot does not always drop to zero, instead, it will rise again at higher frequency and then drop to zero. So there is a peak at higher frequency. Why does that happen? How to deal with it? I will appreciate if anyone can help. Thanks.

Sincerely yours,

Michael Kwan

[Che-Yen Wang]

It general means that you have some information at the higher frequency region or you are refining noise (overfitting). The later may happen if you are running a quite older version of auto3dem and you have manually set your resolution to the Nyquist limit.

1. If you have single defocus dataset, this means you don't have information on certain frequency. You may see the FSC curve drops to zero, raises up and then drops to zero again. By combining multiple defocus dataset, this may disappear.

 

2. If you refine the dataset at very low frequency (say 20 Å), this may suggest you can increase your frequency limit to the higher frequency range (assuming you are manually running auto3dem)

 

3. If you refine your structure to the Nyquist, this may be an indication that you are refining high frequency noise.

--

---
You received this message because you are subscribed to the Google Groups "Auto3DEM" group.
To unsubscribe from this group and stop receiving emails from it, send an email to auto3dem+u...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Giovanni Cardone]

Hi Michael,

as Che-Yen already pointed out, there could be many reasons for this behavior, and more information would be needed to explain your specific case.
In any case I would recommend the following article by Penczek where he nicely explains the shortcomings of the FSC (see last section 'Resolution assessment in practice').

Best,
Giovanni


Penczek PA., "Resolution measures in molecular electron microscopy." Methods Enzymol. 2010;482:73-100. doi: 10.1016/S0076-6879(10)82003-8.

-- 
Giovanni Cardone
Baker's Laboratory
University of California, San Diego
Department of Chemistry and Biochemistry
9500 Gilman Drive, NSB 4107, MC-0378
La Jolla, CA, 92093-0378
(858)534-5852

[Jian Guan]

Thank you so much. The problem is the FSC of all of my datasets have the samll peak at high frequency, which make me think maybe something to do with the setting of Auto3dem. My datasets are with large defocus range (1.22 to 6.6).

Thanks,

Jian

--

---
You received this message because you are subscribed to a topic in the Google Groups "Auto3DEM" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/auto3dem/JhBIOe9QUYw/unsubscribe.
To unsubscribe from this group and all its topics, send an email to auto3dem+u...@googlegroups.com.