VMD-L Mailing List

From: Ryan McGreevy (ryanmcgreevy_at_ks.uiuc.edu)
Date: Mon Feb 23 2015 - 10:50:26 CST

If what you mean by "I always load it into VMD to check fitting and found
no weirdness." is that you load the griddx potential file and check it
visually (not using ccc), then yes doing this will be fine. The density
files and potential files look visually more or less the same. However,
whereas the higher grid values of a density map tend to be closer to the
structure, the higher values of the potential are further away (thus
inverted). This is how the potential works, by driving your structure from
high to low. So visually the maps will look the same, just inverted, which
you can see if you adjust the isovalue in its representation. This is why
taking the cross correlation of the potential instead of the density map
will give you roughly the same cc, but negative. The cc is computed by
creating a simulated density from the structure and comparing it to the
density map. If compared with the potential map, this correlation will be
negative (anticorrelated).

On Mon Feb 23 2015 at 4:18:43 AM Kevin C Chan <cchan2242-c_at_my.cityu.edu.hk>
wrote:

> Thanks for all the replies.
>
> John,
> I know that we could load mrc files and I do it all the time to check
> fittings. I was just not sure whether the plugin mdff check can read mrc.
> Turns out it does.
>
> Ryan and Tristan,
> I understand directly trimming the potential is bad but what I was
> interested in is how my system response to separate MDFF for each dimer in
> vacuum and will soon return to the original 4-dimer big system so I think
> it is fine. Thanks for the reminder anyway.
>
> I have used the very original mrc file instead and it works fine giving
> reasonable ccc. However what I could not understand is why my previous
> calculation gave a negative value (with similar magnitude). As Tristan has
> mentioned, the griddx map might be inverted. However what I have done is
> just trimmed the original potential file (griddx) and I always load it into
> VMD to check fitting and found no weirdness.
> ​So in what manner my griddx is inverted?
>
>
> Thanks in advance,
>
> Kevin
>
> On Wed, Feb 18, 2015 at 6:13 AM, Tristan Croll <tristan.croll_at_qut.edu.au>
> wrote:
>
>> If I understand correctly, your CCC is fine. The griddx map is inverted
>> (zero becomes 1 and vice versa), which means it's now *anti*correlated to
>> your structure. -1 is perfect anti-correlation, +1 is perfect correlation,
>> and 0 is no correlation.
>>
>> Cheers,
>>
>> Tristan
>>
>>
>>
>> Tristan Croll
>> Lecturer
>> Faculty of Health
>> School of Biomedical Sciences
>> Institute of Health and Biomedical Engineering
>> Queensland University of Technology
>> 60 Musk Ave
>> Kelvin Grove QLD 4059 Australia
>> +61 7 3138 6443
>>
>> This email and its attachments (if any) contain confidential information
>> intended for use by the addressee and may be privileged. We do not waive
>> any confidentiality, privilege or copyright associated with the email or
>> the attachments. If you are not the intended addressee, you must not use,
>> transmit, disclose or copy the email or any attachments. If you receive
>> this email by mistake, please notify the sender immediately and delete the
>> original email.
>>
>>
>>
>> On 18 Feb 2015, at 4:05 am, Kevin C Chan <cchan2242-c_at_my.cityu.edu.hk>
>> wrote:
>>
>> Thanks so much for the reply.
>>
>> Your guess is right, I'm in fact using the potential map generated by
>> griddx command. However what was because I have used it for my previous
>> (not yet trimmed) structure and CCC calculations and it seems fine as I got
>> values of 0.7~0.8. I'm confused why the sign reverses, which indicates a
>> rather bad correlation, simply because I trimmed the system (of cause I run
>> new MDFFs for the trimmed ones).
>>
>> I'm willing to try my very original density map instead however it has
>> .mrc as extension and I didn't find it available in the plugin's input
>> list.
>>
>> Thanks in advance,
>>
>> Kevin
>>
>> Sent from my iPhone
>>
>> On 18 Feb, 2015, at 00:19, Ryan McGreevy <ryanmcgreevy_at_ks.uiuc.edu>
>> wrote:
>>
>> The threshold is based on the map simulated from your structure. When
>> you calculate the correlation, are you sure that you are comparing to the
>> density map, and not the potential map (the map you fit to during mdff,
>> created with the griddx command)? In cases of such bad correlation, my
>> first guess is that you are using the potential instead of the density
>> during analysis.
>>
>> On Mon Feb 16 2015 at 9:42:31 PM Kevin C Chan <
>> cchan2242-c_at_my.cityu.edu.hk> wrote:
>>
>>> Dear Users,
>>>
>>> I am now using mdff plugin to analyse my mdff results.
>>>
>>> I am using a quite crude density map (as it is trimmed from a much
>>> bigger one) in which there are much unoccupied density areas. As the
>>> unoccupied areas are few angstroms away from my structure, I hope they
>>> won’t affect much. Therefore I resort to local ccc calculations. I have
>>> chosen threshold of 0.5 (and waiting for results of 0.2) however it still
>>> gives a very bad ccc, say -0.7 to -0.8. I am curious that the threshold was
>>> referred to the given density map or the simulated one from my input
>>> structure? If it is the given one, then the bad ccc values come from the
>>> unoccupied areas in the simulated map, right?
>>>
>>> Thanks in advance,
>>>
>>> Kevin
>>> PhD Candidate
>>> Department of Physics and Material Science
>>> City University of Hong Kong
>>> ukevi_at_gmx.hk
>>>
>>>
>