unixjunkie / pharmer Goto Github PK

Pharmer: Efficient and Exact 3D Pharmacophore Search
Copyright (C) 2011  David Ryan Koes and the University of Pittsburgh

LICENSE

This program is free software; you can redistribute it and/or
modify it under the terms of the GNU General Public License
as published by the Free Software Foundation; either version 2
of the License, or (at your option) any later version.

This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
GNU General Public License for more details.

You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA  02110-1301, USA.


BUILDING

Currently, the easiest way to build pharmer is to checkout the source code
using Eclipse and build it with the automatically generated make files.
The contribution of a standard autoconf/make build system would be greatly
appreciated.

There are several dependencies on third party libraries:
Boost (currently building against version 1.45)
http://www.boost.org/

JsonCpp
http://jsoncpp.sourceforge.net/

FastCGI
http://www.fastcgi.com

cgicc
http://www.gnu.org/software/cgicc/

tcmalloc (google performance tools)
http://code.google.com/p/google-perftools/

OpenBabel
http://openbabel.org/


USING

The --help option will provide the following:

USAGE: pharmer [options] --cmd command [pharma, dbcreate, dbsearch, server]

OPTIONS:
  -bloom-large=<uint>  - Number of bits for coarse discretization fingerprint.
  -bloom-small=<uint>  - Number of bits for fine discretization fingerprint.
  -cmd=<string>        - command [pharma, dbcreate, dbsearch, server]
  -dbdir=<string>      - database directory(s)
  -extra-info          - Output additional molecular properties.  Slower.
  -file-partition      - Partion database slices based on files
  -help                - Display available options (--help-hidden for more)
  -in=<string>         - input file(s)
  -logdir=<string>     - log directory for server
  -max-hits=<n>        - return at most n results
  -max-orient=<n>      - return at most n orientations of each conformation
  -max-rmsd=<number>   - maximum allowed RMSD; default max allowed by query
  -max-weight=<uint>   - maximum allowed molecular weight
  -nthreads=<n>        - utilize n threads; default 1
  -out=<string>        - output file
  -pharmaspec=<string> - pharmacophore specification
  -port=<uint>         - port for server to listen on
  -print               - print results
  -q                   - quiet; suppress informational messages
  -reduceconfs=<n>     - return at most n conformations for each molecule
  -separate-weight     - Segregate database based on molecular weight
  -show-query          - print query points
  -sort-rmsd           - Sort results by RMSD.
  -unweighted-rmsd     - Compute minimal RMSD without radius weights

To IDENTIFY the pharmacophore features of a molecule ex.pdb:
 pharmer pharma -in ex.pdb
 pharmer pharma -in ex.pdb -out out.sdf
 pharmer pharma -in ex.pdb -out out.json
For interactive, graphical editting of pharmacophore features, try
http://zincpharmer.csb.pitt.edu
The saved session file is a json file that can be used as input for Pharmer.

To CREATE a database DB from library.sdf:
 pharmer dbcreate -dbdir DB -in library.sdf
If you have multiple disk drives, you can improve performance by striping the
database across the drives:
 pharmer dbcreate -dbdir /drive1/DB -dbdir /drive2/DB -in library.sdf
If you pre-split the input file things will go faster:
 pharmer dbcreate -dbdir /drive1/DB -dbdir /drive2/DB -file-partition -in library1.sdf -in library2.sdf

To SEARCH a database:
 pharmer dbsearch -dbdir DB -in query.json
 pharmer dbsearch -dbdir DB -in query.sdf
 pharmer dbsearch -dbdir DB -in query.ph4

BUGS (Not Really)

When working with large datasets spread across multiple hard drives,
you will likely run into a number of limits that will result in 
segmentation faults (type 6 or 11).  These changes need to be in place
before building the database, or it will silently create incomplete 
databases that crash when you search them.

Make sure nofile is set high enough (1000000) in /etc/security/limits.conf

Make sure vm.max_map_count is set high enough - add the following to /etc/sysctl.conf
vm.max_map_count=1000000

Make sure you build against libcurl4-openssl-dev, not gnutls, or you
will get fortify_fail errors.

Continuing with curl bugs, DNS timeouts can cause a longjmp error:
http://stackoverflow.com/questions/9191668/error-longjmp-causes-uninitialized-stack-frame
to resolve these rebuild curl with --enable-ares

You can ignore all the curl issues entirely if you build with -DSKIP_REGISTERZINC
which is probably what you want unless you are using the full ZINC database.

Some versions of eigen3 will trigger an assert, OBJECT_ALLOCATED_ON_STACK_IS_TOO_BIG, at compile time.
This can be ignored in the latest versions of eigen3 with -DEIGEN_DISABLE_STACK_SIZE_ASSERT
If your version hasn't had this patch applied (which is currently true with Ubuntu 14.04),
then you will have to modify Eigen/src/Core/DenseStorage.h appropriately.

Currently, on Ubuntu 14.04, g++ 4.8 experiences an internal compiler error.
Use g++-4.6 to get around this.

BUGS (Really)

Send any bug reports to [email protected] with a complete test case.

CITING

Please use the citation from http://pubs.acs.org/doi/abs/10.1021/ci200097m

DEFAULT PHARMACOPHORE DEFINITIONS
Aromatic 18 0 1.1 1 0.1 
a1aaaaa1
a1aaaa1

HydrogenDonor 1 1 0.5 1 0.1 
[#7!H0&!$(N-[SX4](=O)(=O)[CX4](F)(F)F)]
[#8!H0&!$([OH][C,S,P]=O)]
[#16!H0]

HydrogenAcceptor 89 2 0.5 1 0.1 
[#7&!$([nX3])&!$([NX3]-*=[!#6])&!$([NX3]-[a])&!$([NX4])&!$(N=C([C,N])N)]
[$([O])&!$([OX2](C)C=O)&!$(*(~a)~a)]

PositiveIon 7 3 0.75 0 0.1 
[+,+2,+3,+4]
[$(CC)](=N)N
[$(C(N)(N)=N)]
[$(n1cc[nH]c1)]

NegativeIon 8 4 0.75 0 0.1 
[-,-2,-3,-4]
C(=O)[O-,OH,OX1]
[$([S,P](=O)[O-,OH,OX1])]
c1[nH1]nnn1
c1nn[nH1]n1
C(=O)N[OH1,O-,OX1]
C(=O)N[OH1,O-]
CO(=N[OH1,O-])
[$(N-[SX4](=O)(=O)[CX4](F)(F)F)]

Hydrophobic 6 5 1 0 2 
a1aaaaa1
a1aaaa1
[$([CH3X4,CH2X3,CH1X2,F,Cl,Br,I])&!$(**[CH3X4,CH2X3,CH1X2,F,Cl,Br,I])]
[$(*([CH3X4,CH2X3,CH1X2,F,Cl,Br,I])[CH3X4,CH2X3,CH1X2,F,Cl,Br,I])&!$(*([CH3X4,CH2X3,CH1X2,F,Cl,Br,I])([CH3X4,CH2X3,CH1X2,F,Cl,Br,I])[CH3X4,CH2X3,CH1X2,F,Cl,Br,I])]([CH3X4,CH2X3,CH1X2,F,Cl,Br,I])[CH3X4,CH2X3,CH1X2,F,Cl,Br,I]
*([CH3X4,CH2X3,CH1X2,F,Cl,Br,I])([CH3X4,CH2X3,CH1X2,F,Cl,Br,I])[CH3X4,CH2X3,CH1X2,F,Cl,Br,I]
[C&r3]1~[C&r3]~[C&r3]1
[C&r4]1~[C&r4]~[C&r4]~[C&r4]1
[C&r5]1~[C&r5]~[C&r5]~[C&r5]~[C&r5]1
[C&r6]1~[C&r6]~[C&r6]~[C&r6]~[C&r6]~[C&r6]1
[C&r7]1~[C&r7]~[C&r7]~[C&r7]~[C&r7]~[C&r7]~[C&r7]1
[C&r8]1~[C&r8]~[C&r8]~[C&r8]~[C&r8]~[C&r8]~[C&r8]~[C&r8]1
[CH2X4,CH1X3,CH0X2]~[CH3X4,CH2X3,CH1X2,F,Cl,Br,I]
[$([CH2X4,CH1X3,CH0X2]~[$([!#1]);!$([CH2X4,CH1X3,CH0X2])])]~[CH2X4,CH1X3,CH0X2]~[CH2X4,CH1X3,CH0X2]
[$([CH2X4,CH1X3,CH0X2]~[CH2X4,CH1X3,CH0X2]~[$([CH2X4,CH1X3,CH0X2]~[$([!#1]);!$([CH2X4,CH1X3,CH0X2])])])]~[CH2X4,CH1X3,CH0X2]~[CH2X4,CH1X3,CH0X2]~[CH2X4,CH1X3,CH0X2]
[$([S]~[#6])&!$(S~[!#6])]