edit all the values in a specific column based on row numbers range

Question

I have a PDB file (coordinates of atoms in a protein) on a Linux machine:

ATOM      1   N  GLY A   1       0.535  51.766   5.682  1.00  0.00              
ATOM      2  CA  GLY A   1      -0.712  50.962   5.596  1.00  0.00              
ATOM      3   C  GLY A   1      -1.243  50.872   4.179  1.00  0.00              
ATOM      4   O  GLY A   1      -1.313  51.888   3.492  1.00  0.00              
ATOM      5   N  GLN A   2      -1.600  49.664   3.737  1.00  0.00              
ATOM      6  CA  GLN A   2      -2.221  49.468   2.423  1.00  0.00              
ATOM      7   C  GLN A   2      -3.542  48.719   2.507  1.00  0.00              
ATOM      8   O  GLN A   2      -3.722  47.844   3.356  1.00  0.00              
ATOM      9  CB  GLN A   2      -1.280  48.738   1.468  1.00  0.00              
ATOM     10  CG  GLN A   2      -0.976  47.294   1.830  1.00  0.00              
....     ..  ..   .. .   .       ....   ....     ....   ....  ....
TER   SPLIT LINE FOR INTERNAL USE ONLY
ATOM      1  O5'  G  A   1     -44.412  97.503  31.177  1.00  0.00              
ATOM      2  C5'  G  A   1     -45.447  96.803  31.882  1.00  0.00              
ATOM      3  C4'  G  A   1     -45.225  95.295  31.894  1.00  0.00              
ATOM      4  O4'  G  A   1     -46.441  94.578  31.654  1.00  0.00              
ATOM      5  C3'  G  A   1     -44.328  94.850  30.748  1.00  0.00              
ATOM      6  O3'  G  A   1     -42.943  94.877  31.129  1.00  0.00              
ATOM      7  C2'  G  A   1     -44.804  93.425  30.542  1.00  0.00              
ATOM      8  O2'  G  A   1     -44.163  92.592  31.466  1.00  0.00              
ATOM      9  C1'  G  A   1     -46.304  93.444  30.772  1.00  0.00              
ATOM     10  N9   G  A   1     -46.965  93.699  29.495  1.00  0.00
....     ..  ..   .  .   .     .......  ......   .....  ....   ...

The TER record explicitly marks the end of a particular amino acid chain. I want to change the chain ID of the protein at the 5th column by awk to assign the correct ID to the new chain after TER.

Expected Output:

ATOM      1   N  GLY A   1       0.535  51.766   5.682  1.00  0.00              
ATOM      2  CA  GLY A   1      -0.712  50.962   5.596  1.00  0.00              
ATOM      3   C  GLY A   1      -1.243  50.872   4.179  1.00  0.00              
ATOM      4   O  GLY A   1      -1.313  51.888   3.492  1.00  0.00              
ATOM      5   N  GLN A   2      -1.600  49.664   3.737  1.00  0.00              
ATOM      6  CA  GLN A   2      -2.221  49.468   2.423  1.00  0.00              
ATOM      7   C  GLN A   2      -3.542  48.719   2.507  1.00  0.00              
ATOM      8   O  GLN A   2      -3.722  47.844   3.356  1.00  0.00              
ATOM      9  CB  GLN A   2      -1.280  48.738   1.468  1.00  0.00              
ATOM     10  CG  GLN A   2      -0.976  47.294   1.830  1.00  0.00                 
TER   SPLIT LINE FOR INTERNAL USE ONLY
ATOM      1  O5'  G  B   1     -44.412  97.503  31.177  1.00  0.00              
ATOM      2  C5'  G  B   1     -45.447  96.803  31.882  1.00  0.00              
ATOM      3  C4'  G  B   1     -45.225  95.295  31.894  1.00  0.00              
ATOM      4  O4'  G  B   1     -46.441  94.578  31.654  1.00  0.00              
ATOM      5  C3'  G  B   1     -44.328  94.850  30.748  1.00  0.00              
ATOM      6  O3'  G  B   1     -42.943  94.877  31.129  1.00  0.00              
ATOM      7  C2'  G  B   1     -44.804  93.425  30.542  1.00  0.00              
ATOM      8  O2'  G  B   1     -44.163  92.592  31.466  1.00  0.00              
ATOM      9  C1'  G  B   1     -46.304  93.444  30.772  1.00  0.00              
ATOM     10  N9   G  B   1     -46.965  93.699  29.495  1.00  0.00

Everything needs to be separated with the same spaces, this following arrangement would be wrong:

ATOM   3674  CD1 PHE A 460       2.350  79.471  35.466  1.00  0.00              
ATOM   3675  CD2 PHE A 460       1.037  81.443  35.196  1.00  0.00              
ATOM   3676  CE1 PHE A 460       2.425  79.321  34.080  1.00  0.00              
ATOM   3677  CE2 PHE A 460       1.108  81.298  33.805  1.00  0.00              
ATOM   3678  CZ  PHE A 460       1.805  80.232  33.250  1.00  0.00              
TER SPLIT LINE FOR B USE ONLY
ATOM 1 O5' G B 1 -44.412 97.503 31.177 1.00 0.00
ATOM 2 C5' G B 1 -45.447 96.803 31.882 1.00 0.00
ATOM 3 C4' G B 1 -45.225 95.295 31.894 1.00 0.00
ATOM 4 O4' G B 1 -46.441 94.578 31.654 1.00 0.00
ATOM 5 C3' G B 1 -44.328 94.850 30.748 1.00 0.00

In addition, the file ends with this:

TER
ENDMDL

There is a blank line at the end of the file which needs to be left as it is

yes, you are correct, I edited my question. Yes, the white space has to be maintained, that's the thing — Paolo Lorenzini
– Paolo Lorenzini, Commented Apr 17 at 17:25
And what do you want to change A to in the 5th column? To B? Please be explicit. Why are some changed and others not? For example, I see ATOM 3674 CD1 PHE A 460 in your output. Why is that A and not B? And is there one TER or many? Which one should we use? — terdon
– terdon ♦, Commented Apr 17 at 17:26
I only would like to change the 5th column which corresponds to the protein chain ID, the other As need to stay as they are — Paolo Lorenzini
– Paolo Lorenzini, Commented Apr 17 at 17:30
I really don't think PDB files are tab-delimited, by the way. For example, see rcsb.org/structure/9K38. That isn't tab-delimited. Not even the ATOM lines are tab-delimited: sed -n '/\t/p' 9k38.pdb | wc -l returns 0. I think you're just used to seeing the columns aligned, but you likely don't have tabs in your actual file. — terdon
– terdon ♦, Commented Apr 17 at 17:46
you might be right, they are aligned like in a tab but the separator is not /t, it is a "fixed-width columns" separator, but the POSIX works — Paolo Lorenzini
– Paolo Lorenzini, Commented Apr 17 at 17:55

Ed Morton · Accepted Answer · 2025-04-17 17:41:13Z

If your fields are tab-delimited then this might be what you want, using any awk:

$ awk 'BEGIN{FS=OFS="\t"} f && NF{ $5="B"} $1=="TER"{ f=1 } 1' file
ATOM    1       N       GLY     A       1       0.535   51.766  5.682   1.00    0.00
ATOM    2       CA      GLY     A       1       -0.712  50.962  5.596   1.00    0.00
ATOM    3       C       GLY     A       1       -1.243  50.872  4.179   1.00    0.00
ATOM    4       O       GLY     A       1       -1.313  51.888  3.492   1.00    0.00
ATOM    5       N       GLN     A       2       -1.600  49.664  3.737   1.00    0.00
ATOM    6       CA      GLN     A       2       -2.221  49.468  2.423   1.00    0.00
ATOM    7       C       GLN     A       2       -3.542  48.719  2.507   1.00    0.00
ATOM    8       O       GLN     A       2       -3.722  47.844  3.356   1.00    0.00
ATOM    9       CB      GLN     A       2       -1.280  48.738  1.468   1.00    0.00
ATOM    10      CG      GLN     A       2       -0.976  47.294  1.830   1.00    0.00
TER     SPLIT   LINE    FOR     INTERNAL        USE     ONLY
ATOM    1       O5'     G       B       1       -44.412 97.503  31.177  1.00    0.00
ATOM    2       C5'     G       B       1       -45.447 96.803  31.882  1.00    0.00
ATOM    3       C4'     G       B       1       -45.225 95.295  31.894  1.00    0.00
ATOM    4       O4'     G       B       1       -46.441 94.578  31.654  1.00    0.00
ATOM    5       C3'     G       B       1       -44.328 94.850  30.748  1.00    0.00
ATOM    6       O3'     G       B       1       -42.943 94.877  31.129  1.00    0.00
ATOM    7       C2'     G       B       1       -44.804 93.425  30.542  1.00    0.00
ATOM    8       O2'     G       B       1       -44.163 92.592  31.466  1.00    0.00
ATOM    9       C1'     G       B       1       -46.304 93.444  30.772  1.00    0.00
ATOM    10      N9      G       B       1       -46.965 93.699  29.495  1.00    0.00

otherwise this might be what you want, using any POSIX awk and no matter what the white space is between fields as long as you can't have white space within the first 5 fields too:

$ cat tst.sh
#!/usr/bin/env bash

awk '
    f && match($0, /^([^[:space:]]+[[:space:]]+){4}/) {
        $0 = substr($0,1,RLENGTH) "B" substr($0,RSTART+RLENGTH+1)
    }
    $1 == "TER" { f=1 }
    { print }
' "${@:--}"

$ ./tst.sh file
ATOM      1   N  GLY A   1       0.535  51.766   5.682  1.00  0.00
ATOM      2  CA  GLY A   1      -0.712  50.962   5.596  1.00  0.00
ATOM      3   C  GLY A   1      -1.243  50.872   4.179  1.00  0.00
ATOM      4   O  GLY A   1      -1.313  51.888   3.492  1.00  0.00
ATOM      5   N  GLN A   2      -1.600  49.664   3.737  1.00  0.00
ATOM      6  CA  GLN A   2      -2.221  49.468   2.423  1.00  0.00
ATOM      7   C  GLN A   2      -3.542  48.719   2.507  1.00  0.00
ATOM      8   O  GLN A   2      -3.722  47.844   3.356  1.00  0.00
ATOM      9  CB  GLN A   2      -1.280  48.738   1.468  1.00  0.00
ATOM     10  CG  GLN A   2      -0.976  47.294   1.830  1.00  0.00
TER   SPLIT LINE FOR BNTERNAL USE ONLY
ATOM      1  O5'  G  B   1     -44.412  97.503  31.177  1.00  0.00
ATOM      2  C5'  G  B   1     -45.447  96.803  31.882  1.00  0.00
ATOM      3  C4'  G  B   1     -45.225  95.295  31.894  1.00  0.00
ATOM      4  O4'  G  B   1     -46.441  94.578  31.654  1.00  0.00
ATOM      5  C3'  G  B   1     -44.328  94.850  30.748  1.00  0.00
ATOM      6  O3'  G  B   1     -42.943  94.877  31.129  1.00  0.00
ATOM      7  C2'  G  B   1     -44.804  93.425  30.542  1.00  0.00
ATOM      8  O2'  G  B   1     -44.163  92.592  31.466  1.00  0.00
ATOM      9  C1'  G  B   1     -46.304  93.444  30.772  1.00  0.00
ATOM     10  N9   G  B   1     -46.965  93.699  29.495  1.00  0.00

awk 'BEGIN{FS=OFS="\t"} f{ $5="B"} $1=="TER"{ f=1 } 1' file this added a B at the end of the file , but the POSIX worked really well, appreciated — Paolo Lorenzini
– Paolo Lorenzini, Commented Apr 17 at 17:39
@PaoloLorenzini if that script added a B to the end of the output file then you have a blank line at the end of your input file. I just tweaked the awk script to not modify such lines but you should edit your question to mention there could be blank lines in the input that you want to leave as-is. — Ed Morton
– Ed Morton, Commented Apr 17 at 17:41

Romeo Ninov · Accepted Answer · 2025-04-17 16:29:25Z

2

You can try something like:

awk '{if($1=="TER") flag=1;if(flag==1) $5="B"}1' input_file

this code check if the first token is = "TER" and set flag. Then if check if flag is = 1 and if yes set value of token 5 to "B". And on the end print the entire line

You can modify it like this to make it like table with -R for right align of specific columns

awk '{if($1=="TER") flag=1;if(flag==1) $5="B"}1' qq|column -t -R 2,3,7,8,9

answered Apr 17 at 16:29

Romeo Ninov

19.5k5 gold badges34 silver badges48 bronze badges

1

it has changed it but the format has changed as well

Paolo Lorenzini
– Paolo Lorenzini

2025-04-17 17:21:46 +00:00
Commented Apr 17 at 17:21

Add a comment |

cas · Accepted Answer · 2025-04-18 08:22:12Z

If you want to correctly parse and manipulate PDB files, you should use a language which has a library module capable of doing so.

For example, perl and the Bio::PDB::Structure module -- a "Perl module for parsing and manipulating Protein Databank (PDB) files"

From the module's description:

This module combines tools that are commonly used to analyze proteins and nucleic acids from a pdb file stuctures.

The main benefits of using the module are its ability to parse and print out a pdb structure with minimum effort.

However in addition to that it is possible to do structural alignments, RMSD calculations, atom editons, center of mass calculations, molecule editions and so forth.

Both Atom objects and Molecule objects are defined within this module.

In short, using the right tool for the job would make what you want do as simple as:

read and parse the PDB file into a perl data structure using the functions and methods in the module. e.g. start with $m = Bio::PDB::Structure::Molecule->new;, to create a new object called $m to store the PDB data, then $m->read("molecule.pdb",0); to read and parse the file into $m.
make any changes you need to the data structure. e.g. for your question, you'd iterate over each atom in $m, check if it matches your criteria, and make the required changes.
print the data structure to a new PDB file. This will be automatically output with as many spaces and/or tabs as are needed for a correctly-formatted PDB file. e.g. $m->print("new.pdb");

BTW, see also Bioperl. As the web page says: "The Bioperl Project is an international association of users & developers of open source Perl tools for bioinformatics, genomics and life science".

Using (and/or modifying) existing open-source code can save you many hours, days, weeks, or more re-inventing the same or similar tools that many researchers before you have needed to invent.

There's also Biopython, which is basically the same thing for the python language instead of perl. In particular, check out the Bio.PDB package.

And, more generally, there's the Open Bioinformatics Foundation: "The Open Bioinformatics Foundation (OBF) is a non-profit, volunteer-run group that promotes open source software development and Open Science within the biological research community. Membership in the OBF is free and open to anyone who wants to help promote open source or open science in a biological field."

Stack Exchange Network

edit all the values in a specific column based on row numbers range

3 Answers 3

You must log in to answer this question.

Hot Network Questions

edit all the values in a specific column based on row numbers range

3 Answers 3

You must log in to answer this question.

Related

Hot Network Questions