eDDM help

[Gauri Thapa]

In the eDDM tutorial, the input to function dm is from read.all function which creates a list of class pdb. I have created a list of pdb using

a<-read.pdb("Frame0.pdb")

b<-read.pdb("Frame1.pdb")
c<-read.pdb("Frame2.pdb")
d<-read.pdb("Frame3.pdb")
e<-read.pdb("Frame4.pdb")

pdbs<-c(a,b,c,d,e)

and then used the function as

dm<-dm(pdbs)

But then I am getting an error as

Error in UseMethod("dm") :
no applicable method for 'dm' applied to an object of class "list".

How should I read in the pdbs ?

‌

Thank you

‌

[Gauri Thapa]

I did the following instead of pdbaln used in the tutorial,

xyz<-fit.xyz(fixed=a$xyz, mobile=c(b$xyz, c$xyz, d$xyz, e$xyz), fixed.inds=inds$xyz, mobile.inds=inds$xyz)
dm<-dm(xyz)

pc<-pca(dm)
Error in UseMethod("pca") :
no applicable method for 'pca' applied to an object of class "dmat".

Could anyone please help me.

My aim is to calculate eDDM matrix from the trajectory obtained from the md simulation to see changes in the distance between residues during the simulation. The pdbs that I have used are extracted from the trajectory at certain stride value.

Thanks

‌

‌

[Xinqiu Yao]

Hi,

‌

read.all() is not just to create a list of pdb objects. It aligns atoms, coordinates, etc. and saves it in a particular class that will be needed for the use of dm.(). Why did you manually read pdb files instead of using read.all()?

‌

[Gauri Thapa]

Yes I do know that read.all does alignment too but I do not want the sequence alignment of the pdbs to be done as they are from the same protein.

Nonetheless when I do that, I get the following error,

pdbs.aa<-read.all(pdbs)
Error in read.all(pdbs) :
input 'aln' should be a list object as obtained from 'seqaln()', 'pdbaln()', or 'read.fasta.pdb()'
pdbs.aa<-read.all("Frame0.pdb", "Frame1.pdb", "Frame2.pdb", "Frame3.pdb", "Frame4.pdb", "Frame5.pdb")
Error in read.all("Frame0.pdb", "Frame1.pdb", "Frame2.pdb", "Frame3.pdb", :
input 'aln' should be a list object as obtained from 'seqaln()', 'pdbaln()', or 'read.fasta.pdb()'

How should I read multiple pdbs so that I can feed that into “dm” function.

‌

Thanks for your help.

[Xinqiu Yao]

If all pdbs are for the same protein, you can put them in one file separated by “MODEL/ENDMDL”, just like what NMR models are deposited in PDB. Then, a single call of read.pdb(multi=TRUE)`will be sufficient. In this case, use pdb$xyz as the input of dm().

‌

Although PDB format is fine for short trajectories, for long trajectories, we recommend some binary format such as DCD or NetCDF. Then use the bio3d read.dcd() or read.ncdf() function to read the coordinates.

‌

If still use aligned structures, the input should be an alignment object from e.g., seqaln() or pdbaln(), not a LIST of pdb objects by repeatedly reading multiple PDB files or a set of PDB file names. Check ?read.all carefully.

‌

[Gauri Thapa]

Thank you so much Xinqiu Yao. I will definitely try the ENDMDL to create multiple model PDB.

I think I had tried the read.dcd for my trajectory initially but it was taking way too long to finish and give any output for the eddm function.