Sorting biomolecules according to their energy-CodePudding

I have a file which contains 7000 molecules, and their name and energies. Each molecule starts with keyword MODEL 1, second line has energy (-9.102 in below example, first molecule) and 7th line has name of the molecule (S3670 Cefsulodin (sodium).cdx in below example, first molecule). I want to rank/sort all molecules according to their energies such that lowest (most negative) will be first molecule in a resulting text file along with molecule's name. Energy and names could be on same or different lines. I thought to use the grep for parsing but have no experience about sorting according to a value embedded in a sentences. Can somebody please help. Thank you.

MODEL 1
REMARK VINA RESULT:    -9.102      0.000      0.000
REMARK INTER   INTRA:         -13.194
REMARK INTER:                 -12.767
REMARK INTRA:                  -0.427
REMARK UNBOUND:                 0.165
REMARK  Name = S3670 Cefsulodin (sodium).cdx
REMARK  8 active torsions:
REMARK  status: ('A' for Active; 'I' for Inactive)
REMARK    1  A    between atoms: CA_3  and  C_8
REMARK    2  A    between atoms: CA_5  and  N_10
REMARK    3  A    between atoms: C_7  and  C_12
REMARK    4  A    between atoms: C_12  and  N_16
REMARK    5  A    between atoms: C_15  and  C_17
REMARK    6  A    between atoms: C_17  and  C_21
REMARK    7  A    between atoms: C_17  and  S_22
REMARK    8  A    between atoms: C_30  and  C_33
REMARK                            x       y       z     vdW  Elec       q    Type
REMARK                         _______ _______ _______ _____ _____    ______ ____
ROOT
ATOM      1  N   UNL     1      92.970 106.706  73.996  0.00  0.00     0.000 N
ATOM      2  C   UNL     1      93.751 107.062  75.160  0.00  0.00     0.000 C
MODEL 1
REMARK VINA RESULT:    -6.812      0.000      0.000
REMARK INTER   INTRA:         -12.561
REMARK INTER:                 -11.387
REMARK INTRA:                  -1.175
REMARK UNBOUND:                -1.767
REMARK  Name = S3836 6-Gingerol.cdx
REMARK  10 active torsions:
REMARK  status: ('A' for Active; 'I' for Inactive)
REMARK    1  A    between atoms: C_1  and  C_2
REMARK    2  A    between atoms: C_1  and  C_12
REMARK    3  A    between atoms: C_2  and  C_3
REMARK    4  A    between atoms: C_3  and  C_4
REMARK    5  A    between atoms: C_4  and  C_5
REMARK    6  A    between atoms: C_5  and  C_6
REMARK    7  A    between atoms: C_6  and  C_7
REMARK    8  A    between atoms: C_7  and  C_8
REMARK    9  A    between atoms: C_8  and  C_9
REMARK   10  A    between atoms: C_14  and  O_18
REMARK                            x       y       z     vdW  Elec       q    Type
REMARK                         _______ _______ _______ _____ _____    ______ ____
ROOT
ATOM      1  C   UNL     1      89.880 102.122  75.634  0.00  0.00     0.000 C
ENDROOT

CodePudding user response：

You would have to do the following steps:

Split the file at the at "MODEL 1", so that you get 1 model per file
Write a function or script that takes in a single model's data, and prints out a sortable string like ENERGY MODEL_NAME
Apply the function from 2 to every file, to obtain a result where every line is a molecule energy and name
Sort this

All of these are thoroughly explained elsewhere, but here are some common Unix commands you can use:

grep, sed, awk and many others can do this How do I split a string on a delimiter in Bash?
head/tail/cut or sed Getting n-th line of text output and Bash sed/awk/cut to delete nth character
xargs or parallel https://savannah.gnu.org/projects/parallel/
sort https://en.wikipedia.org/wiki/Sort_(Unix)

That said, this is an excellent example of a task that is extremely painful to do in Bash, but a breeze in a better language like Python. Just read the file from STDIN, .split("MODEL 1").splitlines(), and use elementary Python string/list slicing/indexing to pull out your data. You can sort in Python too.

CodePudding user response：

This task can unlikely be completed by a simple command.
An option is retrieving the names and values from original file file first, then combining them and sort the output.
Here is a demonstration using rq (https://github.com/fuyuncat/rquery/releases) to do this.

[ rquery]$ ./rq -i ';' -q "s @row, trim(substr(@raw,strlen('REMARK  Name = '))) | f @raw like 'REMARK  Name*'" samples/biomolecules.txt > /tmp/names.tmp
[ rquery]$ ./rq -i ';' -q "s @row, @4| f @raw like 'REMARK VINA RESULT:*'" samples/biomolecules.txt > /tmp/values.tmp
[ rquery]$ cat /tmp/names.tmp
1;S3670 Cefsulodin (sodium).cdx
2;S3836 6-Gingerol.cdx
[ rquery]$ cat /tmp/values.tmp
1;-9.102
2;-6.812
[ rquery]$ ./rq -q "p d/;/ | m @1,@2 where @fileid=1 | s @r[1][2], @2 | f @fileid=2 and @r[1][1]=@1 | o tofloat(@2)" /tmp/names.tmp /tmp/values.tmp
S3670 Cefsulodin (sodium).cdx   -9.102
S3836 6-Gingerol.cdx    -6.812