
==== Front
Nucleic Acids ResNucleic Acids ResnarnarNucleic Acids Research0305-10481362-4962Oxford University Press 1748884110.1093/nar/gkm276ArticlesPDB2PQR: expanding and upgrading automated preparation of biomolecular structures for molecular simulations Dolinsky Todd J. 1Czodrowski Paul 2Li Hui 3Nielsen Jens E. 4Jensen Jan H. 5Klebe Gerhard 2Baker Nathan A. 1*1Department of Biochemistry and Molecular Biophysics, Center for Computational Biology, Washington University in St. Louis, 700 S. Euclid Ave., Campus Box 8036, St. Louis, MO 63110, USA, 2Department of Pharmaceutical Chemistry, Philipps-University Marburg, Marburg, Germany, 3Department of Chemistry, University of Nebraska-Lincoln, Lincoln, NE, USA, 4School of Biomolecular and Biomedical Science, UCD Conway Institute, University College Dublin, Dublin, Ireland and 5Department of Chemistry, University of Copenhagen, Copenhagen, Denmark*To whom correspondence should be addressed. +1-314-362-2040+1-314-362-0234baker@ccb.wustl.edu7 2007 8 5 2007 8 5 2007 35 Web Server issue W522 W525 31 1 2007 7 4 2007 11 4 2007 © 2007 The Author(s)2007This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/2.0/uk/) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.Real-world observable physical and chemical characteristics are increasingly being calculated from the 3D structures of biomolecules. Methods for calculating pKa values, binding constants of ligands, and changes in protein stability are readily available, but often the limiting step in computational biology is the conversion of PDB structures into formats ready for use with biomolecular simulation software. The continued sophistication and integration of biomolecular simulation methods for systems- and genome-wide studies requires a fast, robust, physically realistic and standardized protocol for preparing macromolecular structures for biophysical algorithms. As described previously, the PDB2PQR web server addresses this need for electrostatic field calculations (Dolinsky et al., Nucleic Acids Research, 32, W665–W667, 2004). Here we report the significantly expanded PDB2PQR that includes the following features: robust standalone command line support, improved pKa estimation via the PROPKA framework, ligand parameterization via PEOE_PB charge methodology, expanded set of force fields and easily incorporated user-defined parameters via XML input files, and improvement of atom addition and optimization code. These features are available through a new web interface (http://pdb2pqr.sourceforge.net/), which offers users a wide range of options for PDB file conversion, modification and parameterization.
==== Body
INTRODUCTION
Due to the importance of electrostatic interactions in biomolecular systems, a variety of computational methods have been developed for evaluating electrostatic forces and energies [see (1–6) and references therein]. Typical computational electrostatics methods for biomolecular systems can be loosely grouped into two categories: ‘explicit solvent’ methods, which treat solvent molecules in full molecular detail, and ‘implicit solvent’ methods, which include solvent–solute interactions in averaged or continuum fashion. Implicit solvent methods are, by definition, limited in detail and therefore lack the atomic-scale accuracy of their explicit solvent counterparts. However, implicit solvent methods have gained increasing popularity, in part due to their elimination of the extensive sampling of solvent configurations required with explicit models (1,3–7).

The basic ingredients of an implicit solvent electrostatics calculation are environmental parameters such as temperature, solvent dielectric and ionic strength; biomolecular atomic coordinates; and parameters for atomic charges and radii. While the environmental parameters are relatively straightforward to specify, the remaining two ingredients can often be difficult to supply. In particular, most biomolecular structures in the Protein Data Bank (PDB) (8) do not contain hydrogen atoms, and many are also missing a fraction of the heavy atom coordinates. The addition of hydrogens and the reconstruction of these missing coordinates is not a trivial process; electrostatic properties obtained from the ‘repaired’ structures can often be very sensitive to the manner in which missing atoms are added and protonation states are assigned (9,10). Furthermore, inconsistent atomic nomenclature and other force field idiosyncrasies can often make the assignment of atomic charges and radii a cumbersome task. An additional obstacle to the use of PDB structures in electrostatics calculations and other biomolecular computational tasks is the accurate assignment of parameters to ‘non-standard’ residues and ligands.

Previously (9), we introduced the freely available PDB2PQR service (http://pdb2pqr.sf.net/), which was designed to facilitate the setup and execution of continuum electrostatics calculations from PDB data, particularly by non-experts. The original PDB2PQR server automated many of the common tasks of preparing structures for continuum electrostatics calculations, including adding a limited number of missing heavy atoms to biomolecular structures, estimating titration states and protonating biomolecules in a manner consistent with favorable hydrogen bonding, assigning charge and radius parameters from a variety of force fields, and finally generating ‘PQR’ output (a PDB-like format with the occupancy and temperature factor columns replaced with charge ‘Q’ and radius ‘R’, respectively) compatible with several popular computational biology electrostatics [APBS (10) and MEAD (11)], docking [AutoDock (12)], simulation [AMBER (13)] and visualization [VMD (14), PyMOL (15) and PMV (16)] packages. Since its inception, we have continued to expand the capabilities of the PDB2PQR server to address the challenges associated with ligand parameterization in PDB files and to include several new features.

METHODS
The PDB2PQR web service is driven by a modular, Python-based collection of routines, which provides considerable flexibility to the software and permits non-interactive, high-throughput usage. The service is available via a number of web mirrors listed at http://pdb2pqr.sf.net/. The source code is also available for download from this link, and due to the portability of Python, PDB2PQR can be executed on a wide range of platforms.

Figure 1 outlines the typical workflow of a PDB2PQR job and summarizes the features described in more detail below. The procedures for reconstruction of missing atoms, hydrogen optimization and APBS input generation were described previously (9) and are essentially unchanged in the current version of the software. Since their initial development, these atom reconstruction options have been greatly improved through a number of bug fixes and code optimization, robust support for separate biomolecular chains, and improved chain termini optimization. The following sections describe modified and new elements of the PDB2PQR pipeline.
Figure 1. Flowchart demonstrating the sequence of operations performed by the pipeline. The process begins with an input PDB file and ends with a parameterized PQR file and, optionally, an APBS input file.



Titration state assignment by PROPKA
Protonation states for titratable protein groups are assigned by PROPKA 1.0 (http://propka.ki.ku.dk) (17). PROPKA utilizes a very fast empirical method to predict pKa values and is successful at predicting unusual pKa values. Recently, a comparative study of several protein pKa prediction methods showed that PROPKA was the most accurate method overall (18). PROPKA uses a heuristic method to compute the pKa perturbations due to desolvation, hydrogen bonding and charge–charge interactions. In the current version of PROPKA, contributions from nucleic acids as well as heteroatoms such as bound ions or ligands to the pKa values are not included. Note that, during the course of titration state assignment, PROPKA generates statistics on residue hydrogen bonding, location and solvent accessibility and Coulombic interactions. This information is available to users as a downloadable text file provided at the end of the PDB2PQR/PROPKA calculation.

Standard residue parameter assignment
PDB2PQR currently allows users to assign protein and (where available) nucleic acid parameters based on explicit solvent AMBER99 (19) and CHARMM27 (20) force fields, the PARSE continuum electrostatics force field (21), a Poisson–Boltzmann-optimized force field by Tan et al. (22), or user-defined force fields. User-defined parameters can be uploaded to the PDB2PQR server in a simple flat-file format described in the PDB2PQR user guide. Additionally, PDB2PQR output can be customized to include a variety of atom naming schemes, including AMBER99 (19), CHARMM22 (20), PARSE (21) and an internal naming scheme based on the IUPAC naming recommendations (23). This flexibility in nomenclature was included to facilitate import of PDB2PQR output into other modeling packages. Additionally, the web server provides a ‘map’ which is output at the end of every PDB2PQR calculation and presents a table of atoms’ name/number, residue name, chain name, AMBER atom type and CHARMM atom type to aid in the interpretation of parameter assignment and the development of user-defined charges and radii.

Ligand parameter assignment
The calculation of ligand charges necessitates detailed information on molecular structure and protonation states due to the large variation in the covalent structures of small-molecule protein ligands. The current version of PDB2PQR therefore requires the ligand structure, protonation state and formal charge to be specified by the user in the popular MOL2 (24) format. Ligand structures in MOL2 format are readily available from popular molecular modeling software and free web services such as PRODRG (25). Future versions of PDB2PQR will include a pdb2mol2 parser and automatic assignment of default ligand protonation states from a small-molecule pKa database.

The calculation of ligand charges in PDB2PQR is based on the partial equalization of orbital electronegativities (PEOE) procedure developed by Gasteiger and Marsili (26). In the PEOE procedure, orbital electronegativities χ are linked to partial atomic charges q by a polynomial expansion (χ = a +b·q + c·q2 + d·q3). The coefficients a, b, c and d were optimized by Gasteiger and Marsili using gas phase data on ionization potentials and electron affinities. We utilize a PEOE algorithm, which has been optimized by Czodrowski et al. to obtain better agreement between theoretical and experimental solvation energies for a set of small molecules including the polar amino acids (27). The resulting PEOE_PB charges have been tested for small-molecule complexes with trypsin, thrombin (28) and HIV protease (29), and have been found to give results that are in agreement with experimental values.

Post-processing
The current version of PDB2PQR supports an ‘extension’ directory for user-defined processing of PDB2PQR output. Such extensions might include alternative naming schemes, identification and parameterization of other molecule types, additional hydrogen bond processing, etc. The web servers listed at (http://pdb2pqr.sf.net/) provide only the default PDB2PQR functionality. However, it is straightforward for users to download the PDB2PQR software and setup their own web servers with additional functionality based on custom extensions.

CONCLUSIONS
We have described a number of new features for the free PDB2PQR web server, a service which helps users prepare molecular structures for further computational work by modeling missing atoms, assigning charges and titration states, and providing a mechanism for assignment of ligand parameters. Readers interested in these tasks might also be interested in other servers, which provide complementary services for biomolecular structure processing (30–32). Planned future developments for PDB2PQR include the construction of a pdb2mol2 parser to allow for the automatic parameterization of non-protein atoms, the correct treatment of protein post-translational modification, and the integration of a Poisson–Boltzmann continuum electrostatics-based pKa calculation algorithm into PDB2PQR. We anticipate that the PDB2PQR service will continue to be a helpful addition to the portfolio of tools available to the structural and computational biology communities.

ACKNOWLEDGEMENTS
N.A.B. and T.J.D. were supported by NIH grant GM069702 and the National Biomedical Computation resource (NIH P41 RR08605); J.H.J. and H.L. were supported by NSF grant MCB 0209941; J.H.J. gratefully acknowledges a Skou Fellowship from the Danish Natural Science Research Council; J.E.N. was supported by a Science Foundation Ireland PIYRA grant (04/YI1/M537); G.K. and P.C. were financially supported by the bilateral CERC3 program of CNRS and DFG (KL 1204/3). The authors would like to thank Andy McCammon for contributions to and support of early versions of the PDB2PQR effort. Funding to pay the Open Access publication charges for this article was provided by NIH grant GM069702.

Conflict of interest statement. None declared.
==== Refs
REFERENCES
1 Baker NA   Improving implicit solvent simulations: a Poisson-centric view Curr. Opin. Struct. Biol 2005 15 137 143 15837170 
2 Darden TA   Becker OM  MacKerell AD Jr  Roux B  Watanabe M   Computational Biochemistry and Biophysics 2001 New York Marcel Dekker Inc 91 114 
3 Roux B   Becker OM  MacKerell AD Jr  Roux B  Watanabe M   Computational Biochemistry and Biophysics 2001 New York Marcel Dekker 133 152 
4 Davis ME  McCammon JA   Electrostatics in biomolecular structure and dynamics Chem. Rev 1990 94 7684 7692 
5 Honig B  Nicholls A   Classical electrostatics in biology and chemistry Science 1995 268 1144 1149 7761829 
6 Warshel A  Sharma PK  Kato M  Parson WW   Modeling electrostatic effects in proteins Biochim. Biophys. Acta Proteins Proteomics 2006 1764 1647 1676 
7 Roux B  Simonson T   Implicit solvent models Biophys. Chem 1999 78 1 20 17030302 
8 Berman HM  Westbrook J  Feng Z  Gilliland G  Bhat TN  Weissig H  Shindyalov IN  Bourne PE   The protein data bank Nucleic Acids Res 2000 28 235 242 10592235 
9 Dolinsky TJ  Nielsen JE  McCammon JA  Baker NA   PDB2PQR: an automated pipeline for the setup, execution, and analysis of Poisson–Boltzmann electrostatics calculations Nucleic Acids Res 2004 32 W665 W667 15215472 
10 Baker NA  Sept D  Joseph S  Holst MJ  McCammon JA   Electrostatics of nanosystems: Application to microtubules and the ribosome Proc. Natl. Acad. Sci. USA 2001 98 10037 10041 11517324 
11 Bashford D   Ishikawa Y  Oldehoeft RR  Reynders JVW  Tholburn M   Scientific Computing in Object-Oriented Parallel Environments 1997 1343 Berlin Springer 233 240 
12 Morris GM  Goodsell DS  Halliday RS  Huey R  Hart WE  Belew RK  Olson AJ   Automated docking using a Lamarckian genetic algorithm and an empirical binding free energy function J. Comput. Chem 1998 19 1639 1662 
13 Case DA  Cheatham TE III  Darden T  Gohlke H  Luo R  Merz KM Jr  Onufriev A  Simmerling C  Wang B    The Amber biomolecular simulation programs J. Comput. Chem 2005 26 1668 1688 16200636 
14 Humphrey W  Dalke A  Schulten K   VMD—visual molecular dynamics J. Mol. Graph 1996 14 33 38 8744570 
15 DeLano WL   Palo Alto, CA, The PyMOL Molecular Graphics System 2002 
16 Sanner MF   Python: a programming language for software integration and development J. Mol. Graph. Mod 1999 17 57 61 
17 Li H  Robertson AD  Jensen JH   Very fast empirical prediction and rationalization of protein pKa values Proteins 2005 61 704 721 16231289 
18 Davies MN  Toseland CP  Moss DS  Flower DR   Benchmarking pKa prediction BMC Biochemistry 2006 7 18 16749919 
19 Wang JM  Cieplak P  Kollman PA   How well does a restrained electrostatic potential (RESP) model perform in calculating conformational energies of organic and biological molecules? J. Comput. Chem 2000 21 1049 1074 
20 MacKerell AD Jr  Bashford D  Bellot M  Dunbrack RL Jr  Evanseck JD  Field MJ  Fischer S  Gao J  Guo H    All-atom empirical potential for molecular modeling and dynamics studies of proteins J. Phys. Chem. B 1998 102 3586 3616 
21 Sitkoff D  Sharp KA  Honig B   Accurate calculation of hydration free energies using macroscopic solvent models J. Phys. Chem 1994 98 1978 1988 
22 Tan C  Yang L  Luo R   How well does Poisson–Boltzmann implicit solvent agree with explicit solvent? A quantitative analysis J. Phys. Chem. B 2006 110 18680 18687 16970499 
23 Markley JL  Bax A  Arata Y  Hilbers CW  Kaptein R  Sykes BD  Wright PE  Wüthrich K   Recommendations for the presentation of NMR structures of proteins and nucleic acids J. Mol. Biol 1998 280 933 952 9671561 
24 SYBYL Molecular Modeling Software 2006 7.2 ed St. Louis, MO Tripos Inc. (http://www.tripos.com/mol2/mol2_format3.html ) 
25 van Aalten DMF  Bywater R  Findlay JBC  Hendlich M  Hooft RWW  Vriend G   PRODRG, a program for generating molecular topologies and unique molecular descriptors from coordinates of small molecules J. Comput. Aided Mol. Des 1996 10 255 262 8808741 
26 Gasteiger J  Marsili M   Iterative partial equalization of orbital electronegativity—a rapid access to atomic charges Tetrahedron 1980 36 3219 3228 
27 Czodrowski P  Dramburg I  Sotriffer CA  Klebe G   Development, validation, and application of adapted PEOE charges to estimate pKa values of functional groups in protein-ligand complexes Proteins 2006 65 424 437 16927370 
28 Czodrowski P  Sotriffer CA  Klebe G   Protonation changes upon ligand binding to trypsin and thrombin: structural interpretation based on pKa calculations and ITC experiments J. Mol. Biol 2007 367 1347 1356 17316681 
29 Czodrowski P  Sotriffer CA  Klebe G   Atypical protonation states in the active site of HIV-1 protease: A computational study J. Chem. Inform. Model in press 
30 Gordon JC  Myers JB  Folta T  Shoja V  Heath LS  Onufriev A   H++: a server for estimating pKas and adding missing hydrogens to macromolecules Nucleic Acids Res 2005 33 W368 W371 15980491 
31 Li X  Jacobson MP  Zhu K  Zhao S  Friesner RA   Assignment of polar states for protein amino acid residues using a interaction cluster decomposition algorithm and its application to high resolution protein structure modeling Proteins 2007 66 824 837 17154422 
32 Vriend G   WHAT IF: a molecular modeling and drug design program J. Mol. Graph 1990 8 52 56 29 2268628

