I suspect that PCA will only give you a good estimate of the helix
direction with a long helix. I've done it (less elegantly perhaps) by
superimposing an ideal helix which has its axis along z, then
extracting angles from the superposition matrix. However I'm afraid
I've not packaged this stuff up into a distributable program
Phil
On 16 Apr 2009, at 10:46, James Stroud wrote:
> On Apr 16, 2009, at 2:29 AM, James Stroud wrote:
>> On Apr 16, 2009, at 1:18 AM, Leiman Petr wrote:
>>> I am not sure if it possible to understand _anything_ in
>>> crystallography if it is not clear how to calculate an angle
>>> between two vectors!
>>
>> More difficult is deciding exactly how to define said vectors in
>> the first place, especially when one is attempting to define said
>> vectors for helices "of different subunit[s] in different
>> orientation[s]".
>
> On second thought it's only slightly less trivial than the angle
> calculation: (1) decide the C-alphas of the helix (e.g. DSSP) (2) do
> PCA on their coordinates (3) the first principle component defines
> the direction of the helix.
|