Format of OFILE in PCA output

[Luke Lin]

Hi,

I came across one question for the PCA output of OFILE. It seems that the x coordinate starts from column 30, instead of 31 as required in PDB format:

https://plumed.github.io/doc-v2.3/user-doc/html/pdbreader.html

Here is one example of PCA output pca-comp.pdb:

ATOM     2  X    RES     1   -23.615   2.802  -0.228  0.01  0.01

As you can see, the x coordinate starts from column 30 instead of 31. Is that correct?

Thanks,

Luke

 

[Giovanni Bussi]

Hi

I think there is a bug at this line:

https://github.com/plumed/plumed2/blob/3ba5a59c1d429bc9bf9840e04046c37ca28fd298/src/reference/ReferenceAtoms.cpp#L56

Can you try if this fixes the problem?

Change this:

ofile.printf("ATOM %4d X RES %4u %8.3f%8.3f%8.3f%6.2f%6.2f\n",

To this:

ofile.printf("ATOM %5d X RES %4u %8.3f%8.3f%8.3f%6.2f%6.2f\n",

Giovanni

--
You received this message because you are subscribed to the Google Groups "PLUMED users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users...@googlegroups.com.
To post to this group, send email to plumed...@googlegroups.com.
Visit this group at https://groups.google.com/group/plumed-users.
To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/4edc20d0-f17d-4596-89ac-71cb464135bf%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Luke Lin]

Hi, 

Thanks. That fixes it. But I think even in the previous case, the PCA calculation should be correct? Because it uses the same format in script to read in and output the PDB?

Best,

Luke

[Giovanni Bussi]

Hi,

I am afraid the answer is no. If I look at the current regtests:

output (analysis/rt-pca/pca-comp.pdb.reference)

REMARK TYPE=OPTIMAL

ATOM     1  X    RES     0    -3.002  -0.303   1.042  0.05  0.05

ATOM     2  X    RES     1    -3.375  -0.131   1.046  0.05  0.05

input (mapping/rt-pca/reference.pdb)

ATOM      2  CH3 ACE     1      12.932 -14.718  -6.016  1.00  1.00

ATOM      5  C   ACE     1      21.312  -9.928  -5.946  1.00  1.00

ATOM      9  CA  ALA     2      19.462 -11.088  -8.986  1.00  1.00

Only the latter has the correct format. This means that PCAVARS uses the common PDB reader implemented in PLUMED, that expects the correct columns, whereas PCA writes with the incorrect columns

Giovanni

Giovanni

To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/b760c5f5-fc32-46b7-9fb3-10cd691b5b1b%40googlegroups.com.

[Luke Lin]

Thanks, Giovanni! It may be too much to ask, but I was using PCARMSD with the PCs calculated from the uncorrected version of Plumed. Do you think the format issue will cause any problem? I am worried about if the reading is incorrect, the X coordinates being read in may lose some minus signs. This is what I observed when I converted the PDBs into Gro files using the Gromacs command trjconv

Thanks,

Luke

[Giovanni Bussi]

I am afraid it will affect what you are doing. The best thing that you can do is to try to compute PCARMSD along a trajectory with the wrong pdb and with the right pdb and see if you obtain different values.

Notice that the problem is fixed now (in 2.3/2.4/master branches) so it will be fixed in the forthcoming releases (due in a few days)

Giovanni

To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/0200fc6f-837a-4541-b162-cfdb3d6434a4%40googlegroups.com.

[Luke Lin]

Thanks, Giovanni. I tested it. It is as what you said. Good to find it out!

Luke