Different results of PCA from Bio3d and gromacs

[Former user]

Hi everyone,

I been trying to figure out how PCA works and since i came across Bio3d which is much more user friendly. I tried to carry out the PCA analysis which i had already performed by gromacs. But i find a difference in the clusters and also the values range too high in bio3d. Please have a look at plot and clarify where am going wrong ?

code used for bio3d-

library(bio3d) dcd <- read.dcd('prt-mu-gdp-gtp.dcd') NATOM = 2696 NFRAME= 511 ISTART= 0 last = 511 nstep = 511 nfile = 511 NSAVE = 1 NDEGF = 0 version 24 |================================================| 100% pdb <- read.pdb('prt-mu-gtp.pdb') ca.inds <- atom.select(pdb, elety = 'CA') xyz <- fit.xyz(fixed=pdb$xyz,mobile=dcd, fixed.inds=ca.inds$xyz,mobile.inds=ca.inds$xyz) pc <- pca.xyz(xyz[, ca.inds$xyz]) plot(pc, col=bwr.colors(nrow(xyz)))

sessionInfo() R version 3.5.1 (2018-07-02) Platform: x86_64-apple-darwin15.6.0 (64-bit) Running under: OS X El Capitan 10.11.6

Matrix products: default BLAS: /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib LAPACK: /opt/local/Library/Frameworks/R.framework/Versions/3.5/Resources/lib/libRlapack.dylib

locale: [1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages: [1] stats graphics grDevices utils datasets [6] methods base

other attached packages: [1] bio3d_2.3-4

loaded via a namespace (and not attached): [1] compiler_3.5.1 parallel_3.5.1 tools_3.5.1
[4] Rcpp_1.0.1 grid_3.5.1

[Xinqiu Yao]

Hi,

‌

I didn’t spot any problem in the codes you printed above. Does your system have a single or multiple chains? Is it possible that there are imaging issues in the trajectory - for example, one chain crosses the box boundary but the other does not - that was taken care by gromacs but not by bio3d? You can test it by generating a properly fitted trajectory by gromacs and then using bio3d to just do PCA and compare. Let me know if it is the case.

[Aiswarya Pawar]

Hi,

Sorry to have not logged earlier.

I have two simulations (single chain) wildtype and mutant and concatenated these two simulations into one, and also taken care of the PBC conditions and converted into dcd file. I have used initial PDB structure from the simulation for the Bio3d. I dont understand the issue here.

Also I felt there would be a problem with concatenation i performed wildtype only simulation still am getting similar plots for Bio3d. Am i missing something here.

[Aiswarya Pawar]

i tried again and realised while reading dcd file, i got an error such that

Warning message:
In dcd.header(trj, verbose) :
Check DCD header data is correct, particulary natom

has this got to do with the high variation in data.

[Aiswarya Pawar]

HI,

I redid the pbc condition and it worked. So basically it was an issue with the periodic image.

Thank you for the correction