Non-standard alternate location indicator

[febos]

Good afternoon Dr. Xiang-Jun Lu

I suddenly met a problem with your program.
In my research i need to divide pdb file with more than one model into different files.
for example pdb file 406D contains 4 models so I divide it into 4 files (I attached them).

When I'm trying to run dssr on them I see a quite interesting behavior: files 406D.pdb1 and 406D.pdb3 are processed well, but on files 406D.pdb2 and 406D.pdb4 I have following output:

Code: [Select]

x3dna-dssr -i=406D.pdb2 -o=out/406D.out2 -long-idstr
File <406D.pdb2> contains NO ATOM/HETATM records
...exiting...

Time used: 00:00:00:00


but as you can see from attachment my files are practically identical
What could it be?

For more examples I've also attached files 1ZEV.pdb1 and 1ZEV.pdb2 (first was processed well, second had the same problem)

Best regards,
Eugene.

[xiangjun]

Hi Eugene,

The issues you experienced with 406D.pdb2 etc are due to the non-standard "Alternate location indicator" at column 17 in PDB format. By default, 3DNA/DSSR treats only the standard variants, "A1 ", i.e., when the column #17 character is either " " (space), "A", or "1". For 406D.pdb2, the alternate location indicator on column #17 is "B", as shown below:

Code: [Select]

ATOM      6  O3'B  C A   1      -3.349  35.281  69.551  1.00  9.66           O 
ATOM      7  C2'B  C A   1      -2.839  34.367  67.344  1.00  9.66           C 
ATOM      8  O2'B  C A   1      -4.232  34.217  67.145  1.00  9.66           O 
ATOM      9  C1'B  C A   1      -2.214  35.046  66.123  1.00  9.66           C 
ATOM     10  N1 B  C A   1      -0.760  34.826  66.025  1.00  8.90           N 
ATOM     11  C2 B  C A   1      -0.276  33.554  65.677  1.00  8.90           C 
ATOM     12  O2 B  C A   1      -1.094  32.628  65.441  1.00  8.90           O 

In this case, you need to run DSSR by specifying the --altloc=B option:

Code: [Select]

x3dna-dssr -i=406D.pdb2 --altloc=B
The same rule applies to other entries you listed. Of course, you need to specify the altloc character for each case specifically. Note that for 406D.pdb3, you should add the --altloc=C option as well to get the desired result. The " " (space) altloc is always checked.

It is interesting to note that PDB is handling x-ray crystal structures with MODEL/ENDMDL, and altloc this way. The REMARK 200 in the PDB entry 406d may be helpful.

Code: [Select]

REMARK 200 REMARK: THE CRYSTAL PACKING OF 17-BP RNA DUPLEX IS STATICALLY       
REMARK 200  DISORDERED. AS A RESULT, THE ASYMMETRIC UNIT OF THE CRYSTAL         
REMARK 200  CONTAINS FOUR SUPERIMPOSED ORIENTATIONS OF THE DUPLEX THAT ARE     
REMARK 200  OUT OF REGISTER, SUCH THAT BACKBONES SUPERIMPOSE, BUT BASE         
REMARK 200  IDENTITY DIFFERS.
HTH,

Xiang-Jun

[febos]

Thank you Dr. Lu

[febos]

Good afternoon Dr. Xiang-Jun Lu

I want to report some unusual behaviour of DSSR (latest version - Apr 19)

If I run DSSR without option --altloc=smth on pdb-file in which this option is needed - the console ouput looks something like " no ATOM/HETATM entries found. Time used 00:00:00" and out-file isn't created. But: in the latest version in this case my script gets from DSSR returnvalue=0 (there was not any error). In previous versions (particularly jan 25) my script got returnvalue=1 in such cases (program finished with error).

I need to get returnvalue=1 in such cases so my script can identify when --altloc is needed.

Could you please fix this?

Best regards,
Eugene.

[xiangjun]

Hi Eugene,

The difference you noticed is due to an change in exit code when "no ATOM/HETATM records found". Previously, DSSR used exit(EXIT_FAILURE);, but recently set to exit(EXIT_SUCCESS); because it is more logical to me. This exit could be triggered as in your case, or by specifying a non-PBB/mmCIF file.

Instead of relying on DSSR's exit code in this special case, you could (easily) modify your script to fit your needs.

Xiang-Jun