rubi-za / mode-task Goto Github PK

Hello,
I would like to ask if there is a possible way to exclude several residues from C terminal of a protein while doing PCA. Those residues have high fluctuation that I don't want to take into account in order to not mask any other movement. If there isn't a way, I'll just remove them from trajectory.

I followed the installation steps mentioned in the documents. However still getting the following error

>   File "/home/sbhakat/bin/pca.py", line 73
>     print 'ERROR: Missing trajectory argument.... :(  \nPlease see the help by running \n\nsystem_setup.py -h\n\n '
>                                                                                                                   ^
> SyntaxError: invalid syntax

Any help will be appreciated?

Hej,

So I tried the tool. It produces the files as mentioned however I didn't find some post processing options.
We will get different clusters from a tSNE output. Is there any functionality to dump a PDB corresponding to the cluster centers in the programme. I mean if we do the tSNE but can't say anything about the tSNE clusters or vectors then I guess the method is not much of a use as Gromacs already has the full fledged functionality for PCA and other basic dimensionality reduction techniques. Please let me know if there is anything in the pipeline to integrate post-processing of tSNE?

/SB

Dear team-members,
My congratulations for this wonderful tool. I see that the get_kmo function is taking covariance matrix as the input. Is it nt supposed to take the correlation matrix as the input ?.

I quickly looked into the book "Comparitive approach to using R and python for statistical data analysis" (page 174) referred inside your code from which i get that the input for the KMO is actually a correlation matrix and not the covariance matrix. Just for curiosity, i cross-checked the implementation of KMO score in another library (https://factor-analyzer.readthedocs.io/en/latest/_modules/factor_analyzer/factor_analyzer.html) and i find that they use the correlation matrix indeed.

Am i missing something ?

Thanks,
Bala