Integrating normal mode analysis and correlation network

[leleonp]

Hi,

I have a question and a problem when I try to do apply this vignette. I have a great number of PDB structures modelled previously by me. I could do all the eNMA analysis. Right now i'm trying to make the correlation network analysis, Question: 1. How I can make the correlation network is based on the grps obtained by PCA analysis? In your example, you can make this for GTP and GDI structures, is possible to do this with the grps? 2. After applying the pdbs2pdb and layout.cna functions, I have the following response: Error en layout.cna(net, ref.pdb, full = TRUE, k = 3) : Input 'pdb' is not of class 'pdb' as obtained from 'read.pdb()'

Previously I sent a problem called "problem pdbs2pdb" which is now resolved, but maybe it could be related. Thanks in advance

[Lars Skjærven]

1. I think so. Can you replace the fac argument with your grps when calling filter.dccm ?

2. What is the content of your ref.pdb object? i.e. do:

print(ref.pdb)

[leleonp]

1. I try to do that, but maybe (and probably) my code was wrong

cij <- filter.dccm(cijs0,xyz=pdbs, fac=grps.rd,cutoff.cij=0.35,
                   dcut=10, scut=0, pcut=0.75, ncore=7)

2.I write to you an email with the data attached, because The ref.pdb file has a lot of pages Thanks in advance

[Barry Grant]

Lets try to keep the discussion here folks. You don't have to paste all the output of your print() call, just the start. Basically, and Lars will expand on this, if your print() call does not return something like the below then you have not got a proper pdb class object. Which is consistent with the error you mention.

> pdb

 Call:  read.pdb(file = "1bg2")

   Total Models#: 1
     Total Atoms#: 2733,  XYZs#: 8199  Chains#: 1  (values: A)

     Protein Atoms#: 2527  (residues/Calpha atoms#: 323)
     Nucleic acid Atoms#: 0  (residues/phosphate atoms#: 0)

     Non-protein/nucleic Atoms#: 206  (residues: 174)
     Non-protein/nucleic resid values: [ACT (3), ADP (1), HOH (169), MG (1) ]

   Protein sequence:
      DLAECNIKVMCRFRPLNESEVNRGDKYIAKFQGEDTVVIASKPYAFDRVFQSSTSQEQVY
      NDCAKKIVKDVLEGYNGTIFAYGQTSSGKTHTMEGKLHDPEGMGIIPRIVQDIFNYIYSM
      DENLEFHIKVSYFEIYLDKIRDLLDVSKTNLSVHEDKNRVPYVKGCTERFVCSPDEVMDT
      IDEGKSNRHVAVTNMNEHSSRSHSIFLINVKQENTQTEQKLSGKL...<cut>...AKTI

+ attr: atom, helix, sheet, seqres, xyz,
        calpha, call

Therefore, more information on where you got your ref.pdb object from is required for us to dig in further. Thanks!

[leleonp]

Ok, no problem With the command print(ref.pdb) i get the next message:

$A461T

 Call:  read.pdb(file = fname)

   Total Models#: 1
     Total Atoms#: 593,  XYZs#: 1779  Chains#: 1  (values: A)

     Protein Atoms#: 593  (residues/Calpha atoms#: 593)
     Nucleic acid Atoms#: 0  (residues/phosphate atoms#: 0)

     Non-protein/nucleic Atoms#: 0  (residues: 0)
     Non-protein/nucleic resid values: [ () ]

   Protein sequence:
      MTSRRWFHPNITGVEAENLLLTRGVDGSFLARPSKSNPGDFTLSVRRNGAVTHIKIQNTG
      DYYDLYGGEKFATLAELVQYYMEHHGQLKEKNGDVIELKYPLNCADPTSERWFHGHLSGK
      EAEKLLTEKGKHGSFLVRESQSHPGDFVLSVRTGDDKGESNDGKSKVTHVMIRCQELKYD
      VGGGERFDSLTDLVEHYKKNPMVETLGTVLQLKQPLNTTRINAAE...<cut>...KSFR

+ attr: atom, helix, sheet, seqres, xyz,
        calpha, call

thanks

[Lars Skjærven]

I think your ref.pdb is a list and not a pdb object(?). try with pdb.ref$A461T instead. Note that pdbs2pdb() returns a list of pdb objects.

What is the output of filter.dccm() now? Perhaps fac should be a character vector, so try with as.character(grps) instead.

[leleonp]

Both answer work well. Thanks a lot Best regards

[Lars Skjærven]

ops