Of course, because in the shells that I use it will attempt to do variable name substitution in strings that are double-quoted. (I make no warranties about all possible shells). However if you use single quotes: egrep -v '^ATOM|HETATM.*H$' your.pdb > your_noH.pdb Should work just fine in tcsh, csh at the very least. Phil On 2/6/14 2:52 PM, Pavel Afonine wrote:

Hi Tim,

On 2/6/14, 10:52 AM, Tim Gruene wrote:

...

the simple and qucik command

egrep -v "^ATOM|HETATM.*H$" your.pdb > your_noH.pdb

should also work.

just out of curiosity I did (copy-paste of your example)

egrep -v "^ATOM|HETATM.*H$\" m.pdb > m_noH.pdb

and I got:

Illegal variable name.

Pavel

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

So I guess your contribution at this point in the thread is just to be as difficult as possible ? I dare say if you use it on PROTIN format it won't work either. Try it on a file that actually puts out something that conforms to the PDB standard with the element line at the end, not something that almost conforms to the standard, has non-distinct index numbers, apparently missing spaces on the GLY:N atom. I would hope, modulo the usual list of bugs, that phenix.refine actually writes out something more closely resembling the correct format, in which case Tim's regular expression would actually work. From the original post:

...

Actually i refine my structure with phenix along all hydrogen now

Phil Jeffrey Princeton On 2/6/14 3:09 PM, Pavel Afonine wrote:

Thanks Phil,

did this:

egrep -v '^ATOM|HETATM.*H$' m.pdb > m_noH.pdb

Result:

in input file (m.pdb) I have:

ATOM 1 N GLY A 1 0.504 -0.494 0.924 1.00 7.85 ATOM 2 CA GLY A 1 1.272 0.589 0.277 1.00 6.79 ATOM 3 C GLY A 1 1.700 1.614 1.301 1.00 5.59 ATOM 4 O GLY A 1 1.434 1.460 2.496 1.00 6.04 ATOM 0 H1 GLY A 1 0.452 -1.280 0.308 1.00 7.85 ATOM 0 H2 GLY A 1 0.959 -0.765 1.772 1.00 7.85 ATOM 0 H3 GLY A 1 -0.420 -0.171 1.131 1.00 7.85 ATOM 0 HA2 GLY A 1 2.157 0.171 -0.225 1.00 6.79 ATOM 0 HA3 GLY A 1 0.659 1.070 -0.499 1.00 6.79 END

Output file (m_noH.pdb) contains only:

END

Pavel

On 2/6/14, 12:03 PM, Phil Jeffrey wrote:

...

Of course, because in the shells that I use it will attempt to do variable name substitution in strings that are double-quoted. (I make no warranties about all possible shells). However if you use single quotes:

egrep -v '^ATOM|HETATM.*H$' your.pdb > your_noH.pdb

Should work just fine in tcsh, csh at the very least.

Phil

On 2/6/14 2:52 PM, Pavel Afonine wrote:

...

Hi Tim,

On 2/6/14, 10:52 AM, Tim Gruene wrote:

...

the simple and qucik command

egrep -v "^ATOM|HETATM.*H$" your.pdb > your_noH.pdb

should also work.

just out of curiosity I did (copy-paste of your example)

egrep -v "^ATOM|HETATM.*H$\" m.pdb > m_noH.pdb

and I got:

Illegal variable name.

Pavel

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

For old or new-style pdb files, try: awk '$1!~/ATOM|HETATM/ || $3!~/^H/' old.pdb > new.pdb But I understand some new H-atom names don't start with H, then this won't work. And if the number of atoms is large enough to fill the second field and fuse with HETATM in the first field, it won't work unless you define fixed field widths: awk '$1!~/ATOM|HETATM/ || $3!~/^H/' \ FIELDWIDTHS="6 5 5 4 2 4 4 8 8 8 6 6" \ old.pdb > new.pdb . Maybe a specialized tool like phenix.refine is safest! Pavel Afonine wrote:

Hi Phil,

ok, now I tried this:

egrep -v '^ATOM|HETATM.*H$' m.pdb > m_noH.pdb

Input (m.pdb - file right from PDB):

ATOM 1 N GLY A 1 0.504 -0.494 0.924 1.00 7.85 N ATOM 2 CA GLY A 1 1.272 0.589 0.277 1.00 6.79 C ATOM 3 C GLY A 1 1.700 1.614 1.301 1.00 5.59 C ATOM 4 O GLY A 1 1.434 1.460 2.496 1.00 6.04 O ATOM 0 H1 GLY A 1 0.408 -1.171 0.354 1.00 7.85 H ATOM 0 H2 GLY A 1 0.939 -0.775 1.648 1.00 7.85 H ATOM 0 H3 GLY A 1 -0.298 -0.189 1.160 1.00 7.85 H ATOM 0 HA2 GLY A 1 2.052 0.220 -0.166 1.00 6.79 H ATOM 0 HA3 GLY A 1 0.731 1.013 -0.407 1.00 6.79 H END

Output (m_noH.pdb):

END

Just to be clear: none of the two commands suggested so far worked on a valid PDB file (above). So I thought it might be useful to point this out.

All the best, Pavel

On 2/6/14, 12:21 PM, Phil Jeffrey wrote:

...

So I guess your contribution at this point in the thread is just to be as difficult as possible ? I dare say if you use it on PROTIN format it won't work either.

Try it on a file that actually puts out something that conforms to the PDB standard with the element line at the end, not something that almost conforms to the standard, has non-distinct index numbers, apparently missing spaces on the GLY:N atom.

I would hope, modulo the usual list of bugs, that phenix.refine actually writes out something more closely resembling the correct format, in which case Tim's regular expression would actually work.

From the original post:

...

...

Actually i refine my structure with phenix along all hydrogen now

Phil Jeffrey Princeton

On 2/6/14 3:09 PM, Pavel Afonine wrote:

...

Thanks Phil,

did this:

egrep -v '^ATOM|HETATM.*H$' m.pdb > m_noH.pdb

Result:

in input file (m.pdb) I have:

ATOM 1 N GLY A 1 0.504 -0.494 0.924 1.00 7.85 ATOM 2 CA GLY A 1 1.272 0.589 0.277 1.00 6.79 ATOM 3 C GLY A 1 1.700 1.614 1.301 1.00 5.59 ATOM 4 O GLY A 1 1.434 1.460 2.496 1.00 6.04 ATOM 0 H1 GLY A 1 0.452 -1.280 0.308 1.00 7.85 ATOM 0 H2 GLY A 1 0.959 -0.765 1.772 1.00 7.85 ATOM 0 H3 GLY A 1 -0.420 -0.171 1.131 1.00 7.85 ATOM 0 HA2 GLY A 1 2.157 0.171 -0.225 1.00 6.79 ATOM 0 HA3 GLY A 1 0.659 1.070 -0.499 1.00 6.79 END

Output file (m_noH.pdb) contains only:

END

Pavel

On 2/6/14, 12:03 PM, Phil Jeffrey wrote:

...

Of course, because in the shells that I use it will attempt to do variable name substitution in strings that are double-quoted. (I make no warranties about all possible shells). However if you use single quotes:

egrep -v '^ATOM|HETATM.*H$' your.pdb > your_noH.pdb

Should work just fine in tcsh, csh at the very least.

Phil

On 2/6/14 2:52 PM, Pavel Afonine wrote:

...

Hi Tim,

On 2/6/14, 10:52 AM, Tim Gruene wrote:

...

the simple and qucik command

egrep -v "^ATOM|HETATM.*H$" your.pdb > your_noH.pdb

should also work.

just out of curiosity I did (copy-paste of your example)

egrep -v "^ATOM|HETATM.*H$\" m.pdb > m_noH.pdb

and I got:

Illegal variable name.

Pavel

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi Nat, On 02/06/2014 10:26 PM, Nathaniel Echols wrote:

On Thu, Feb 6, 2014 at 1:07 PM, Edward A. Berry wrote:

...

Maybe a specialized tool like phenix.refine is safest!

At the risk of sounding like a broken record again: Unix shell tricks are great *if you already know how to use these tools* and have a detailed knowledge of the PDB format specification. For everyone else, I do not recommend trying to edit models with anything other than a tool specifically designed for this purpose. This is going to be especially important as the field slowly migrates to mmCIF.

some time ago (about 6 months), Eugene Krissinel sported a similar argument together with the statement that if a program does not exist to do a particular job, it was the developers' task to fix it. At that time I described a problem a student of mine had in data mining, but I never received a response. I am glad somebody else here wrote a little script to sort out the problem, because the numbers were important. With the warning above and the implicit concept to wait for developers, I still would not know. So I am happy about the availability and possibility to use scripting to push my research forward and not having to wait for someone to adjust my ideas to some API. Cheers, Tim P.S.: I checked the command on some output from phenix and noticed that the element type correctly is the last character of the line. Some (older) PDB files from the PDB show two blank spaces, the the regular expression does not work.

For the record, I had to add parentheses to get the egrep command to work on my Mac:

egrep -v '^(ATOM|HETATM).*H$' m.pdb

which does the job. But I do not think this solution is preferable to using phenix.reduce or phenix.pdbtools (or any number of other utilities).

A final note: in my experience, if you deposit a file containing hydrogens to the PDB, they're just going to delete them for you (whether you want this or not!), so it is not necessary to do any additional preparation. In fact, I just did this last week, so we'll see what happens.

-Nat

_______________________________________________ phenixbb mailing list [email protected] http://phenix-online.org/mailman/listinfo/phenixbb

- -- - -- Dr Tim Gruene Institut fuer anorganische Chemie Tammannstr. 4 D-37077 Goettingen GPG Key ID = A46BEE1A -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.12 (GNU/Linux) Comment: Using GnuPG with Icedove - http://www.enigmail.net/ iD8DBQFS9KRQUxlJ7aRr7hoRAsSKAKDaaA2ifkk/9XBsRvuX0mjocFt5lACfSrR6 rek3irmtFwofY5GDSoKsdig= =9nL0 -----END PGP SIGNATURE-----