Home

Version 2.0 User`s Manual

Contents

1. Input Option Usage Default value Required i lt file name gt Input PDB file to define the model none Optional t lt Folder Name gt Folder containing templates of none molecular and interaction defini tions AA Use the default all atom model N A CA Use the default Cy model N A tCG lt Folder Name gt Folder containing templates used for none coarse graining Only necessary when CG enabled c lt string gt Input contact file name none g lt string gt Output gro file name smog gro o lt string gt Output top file name smog top s lt string gt Output contacts file name smog contacts n lt string gt Output ndx file name smog ndx dname lt string gt Default name to use for all output none files backup yes no Enable disable generation of backed none up outputs warnonly Report fatal errors as warnings N A limitbondlength If bond length exceeds limits set it N A to the limiting value limitcontactlength If contact length exceeds limits set N A it to the limiting value TABLE 3 1 Flags supported by SMOG v 2 0 chainNum i3 atomNum_i3 chainNum_j3 atomNum_j3 opt distance etc which should be formatted as a single line per contact whitespace delimited where each line has the two atoms interacting and their respective chain numbers The chains are numbered starting from 1 by the order of occurence in the PDB file The atomNum should
2. lt bond energyGroup bb_n gt lt bond gt lt atom gt C2 lt atom gt lt atom gt C1 lt atom gt lt bond energyGroup bb_n gt lt bond gt lt FUNCTIONAL GROUP gt lt atom gt C1 lt atom gt lt atom gt 04 lt atom gt lt bond energyGroup sc_n gt lt atom gt C1 lt atom gt lt atom gt N9 lt atom gt lt bond gt lt bond energyGroup pr_n gt lt atom gt N9 lt atom gt lt atom gt C8 lt atom gt lt bond gt lt bond energyGroup pr_n gt lt atom gt C8 lt atom gt lt atom gt N7 lt atom gt lt bond gt lt bond energyGroup pr_n gt lt atom gt N7 lt atom gt lt atom gt C5 lt atom gt lt bond gt lt bond energyGroup pr_n gt lt atom gt C5 lt atom gt lt atom gt C6 lt atom gt lt bond gt lt bond energyGroup pr_n gt lt atom gt C6 lt atom gt lt atom gt N6 lt atom gt lt bond gt lt bond energyGroup pr_n gt lt atom gt C6 lt atom gt lt atom gt N1 lt atom gt lt bond gt lt bond energyGroup pr_n gt lt atom gt N1 lt atom gt lt atom gt C2 lt atom gt lt bond gt lt bond energyGroup pr_n gt lt atom gt C2 lt atom gt lt atom gt N3 lt atom gt lt bond gt lt bond energyGroup pr_n gt lt atom gt N3 lt atom gt lt atom gt C4 lt atom gt lt bond gt lt bond energyGroup pr_n gt lt atom gt C4 lt atom gt lt atom gt C5 lt atom gt lt bond gt 3075 3076 3077 3078 3079 3080 3081
3. BONDS gt lt bonds gt lt bond func bond_harmonic 10000 gt lt bType gt lt bType gt lt bType gt lt bType gt lt bond gt lt bond func bond_type6 200 gt lt bType gt lt bType gt lt bType gt MG lt bType gt lt bond gt lt bonds gt LISTING 4 5 Bonds section of b file Recall that each atom is given a bType when they are declared in the bif file Given a particular bonded interaction bonds angles dihedrals impropers the functional form for a bonded interaction is assigned by matching the combination of the bTypes in that interaction The first lt bond gt tag line 3 in Listing 4 5 is used assign a function called bond_harmonic to two bonded atoms of any type Since the vanilla model has only B_1 atoms the s could be replaced with B_1 and the same bonds would be assigned Recall under Listing 4 3 line 2 we defined bond_harmonic as a type 1 function under the bonds directive harmonic bond function The input parameters for bond harmonic are ro and tond In this case we use the special character to tell SMOG to calculate the native bond length ro from the PDB structure file You can instead also give an empirical value for the bond distance This feature is useful when adding nonspecific empirical terms e g an AMBER CHARMM backbone to the Hamiltonian MESE The bType attribute here can take either an exact bond type or a special wildcard character that matches
4. e 1 2 What does SMOG v2 do ee ee o 2 Installation 24 Prerequisites 2 02204 86 es bee a eee Se OR ee eS 22 rt o on Sa ee ae See a EN oA Se Ee ee 2 3 Verify SMOG is properly configured o e 3 Using SMOG v2 al Preparing the input PDB file c0c c 0 0 0243 SLI POB file format co dia a ae eee e a 31 2 Preprocessing 2 2 4 es bea ee Re Gea RR ee 3 2 Generating a Structure Based Model 00004 3 2 1 Default All Atom Model 2 02200 322 Default Cy model occ aa a a Go AUG OB e dad RAR 3 3 1 User provided contact map 3 4 Performing a simulation in Gromacs NAMD 341 GCromas4daod o roda naa a aw a a a es S411 All Atom Model ecos esa e eb eke eae EES 3412 CeModel parsnip ek eo a he Ee a S410 Example s so oos be ee eee eres eee 3 2 Notes amd Hims oo os Sa oe Foe Pak ee a a 3 4 2 1 Domain Decomposition OA GOMSE eoe ge olen Pe a a a a ae eR ee ae a 343 1 Examples os ccce bbe bea eR eR EE ee GAA NAM asidero SR a Se ee eG ee a 4 Template Based Approach 4 1 Introduction to templates eo s ss eosa bs Seba Se ee aa ew e 4 2 SMOG v2 Templates 2 6 2 4258 646 bee Ree daa 4 2 1 Biomolecular Information File bif 4 2 2 Setting Information File Lellb os occiso e eS 4 2 3 Bond Information File b amp Nonbond Information File nb 5 Adding a new residue S
5. pairType P_1 gt S10 lt atom gt pairType P_1 gt C11 lt atom gt pairType P_1 gt C12 lt atom gt pairType P_1 gt C13 lt atom gt pairType P_1 gt C14 lt atom gt pairType P_1 gt C15 lt atom gt pairType P_1 gt C16 lt atom gt LISTING 5 2 Adding the atoms section to the residue structure In this example as in the default models all bonded interactions bonds angles and dihedrals are defined the same for all atoms Therefore only one atom group needs to be defined The bond type bType B_1 is identical for all atoms The contact interactions 2980 2981 2982 2983 2984 2985 2986 2987 2988 2989 2990 2991 2992 2993 2994 2995 2996 2997 2998 2999 3000 3001 3002 3003 3004 3005 3006 3007 3008 3009 3010 3011 3012 3013 3014 3015 3016 3017 3018 3019 Chapter 5 Adding a new residue definition 30 pair Type P_1 are also defined to be the same for all atoms However changing the mass of a specific atom such as sulfur S10 in our example will produce a different non bonded interaction due to a different volume exclusion term The new group type of nbType NB_2 will be defined in the nb file 5 3 3 List all of the atom bonds The chemical structure should tell you how atoms are connected to each other via chemical bonds Inspect those bonds and add them to the lt bonds gt section The lt bonds gt tag encapsulates all the bonds in the biomolecule lt bonds gt lt BA
6. q 1 Note that relative to the original implementation of this model e is decreased by a factor of two and eg is increased As discussed in the SMOG v2 publication in preparation this allows for a longer timestep of 0 002 to be utilized which is larger than the originally implemented 0 0005 When assigning dihedral interaction weights egg and esc dihedrals are first grouped if they have a common middle bond For example in a protein backbone there are up to four dihedral angles that may be defined that have the C Ca bond as the middle bond Each dihedral group is given a summed weight of egg or esc The ratio Rgg sc is set to 1 for nucleic acid dihedral angles and and 2 for protein dihedral angles egg for protein and nucleic acids are equal Dihedral strengths and contact strengths are scaled such that 2 Nec Ec p Y eBB ESC Nec Depp Cesc Neo The sums are over all dihedral angles in the system and N is the number of atoms in the system We are going to add a description here for where to look in the template files for each of these parameters As we continue to work on the manual these details will be provided A 2 The C model The Ca model coarse grains the protein as single bead of unit mass per residue located at the position of the a carbon xg denotes the coordinates of the native state and any subscript 0 signifies a value taken from the native state The potential is giv
7. rcoulomb 1 2 cut off distance for coulomb interactions rvdw 1 2 cut off distance for Vdw interactions pbc no Periodic boundary conditions in all the directions table extension 10 nm Should equals half of the box s longest diagonal tc grps system Temperature coupling tau_t 1 0 Temperature coupling time constant Smaller values stronger coupling ref_t 120 0 In reduced units see Gromacs manual for details Pcoupl no Pressure coupling gen_vel yes Velocity generation gen_temp 50 0 gen_seed 1 ld_seed 1 comm_mode angular center of mass velocity removal LISTING 3 1 Sample mdp file for all atom SMOG models used for Gromacs v4 5 4 6 After you have generated the tpr file with grompp you will need to perform the simu lation To run the simulation issue the command gt mdrun s run tpr noddcheck Is is highly recommended that you explore all Gromacs options in order to ensure max imum performance e g the number of threads being used SMOG model specific requirement To use domain decomposition when performing a simulation in parallel using either threads or MPI you should add the additional flag noddcheck Note that for protein folding you will probably want to avoid domain decomposition and instead use particle decomposition by adding the option pd when on a single node 3 4 1 2 Co Model To run a simulation with the C model the steps are the same as for the
8. 6 and the Ca 5 models are provided as defaults See Appendix A for full details of the default models By running SMOG you will generate the top gro and ndx files necessary to perform a structure based simulation in Gromacs or NAMD Additional output files are provided for your information 3 2 1 Default All Atom Model The all atom potential energy function is defined through the template files found in the directory SMOG_PATH SBM_AA These files define 1 the covalent geometry of amino acids nucleic acids some ligands as well as bioinor ganic atoms 2 the energetic and system parameters e g mass charge interaction strengths To generate all atom forcefield and coordinate files for the default model i e top and gro files issue the command gt smog2 i yourFile pdb AA where yourFile pdb is the name of the file containing your molecular system If you would like to specify a different all atom model then use the command gt smog2 i yourFile pdb t templateDirName where templateDirName is the name of the directory containing the desired template files Chapter 3 Usage 9 3 2 2 Default C model To generate forcefield and coordinate files for the default Ca model issue the command gt smog2 i yourFile pdb CA If you would like to use a different set of CG templates gt smog2 i yourFile pdb t templateDirName tCG CGtemplateDirName Note that an additional set of templates are r
9. improper dihedral angles and planar dihedral angles remain about the values found in file pdb Non bonded atom pairs that are in contact in the provided structure between residues and 7 where i gt j 3 for proteins and i j for RNA are given an attractive 6 12 potential The minimum of each 6 12 interaction is set to the distance of that atom pair in the provided structure All non native interactions between atoms that are not in contact in file pdb are repulsive Contacts were defined according to the Shadow algorithm See Appendix B with an all atom cutoff distance of 6 A and a shadowing radius of 1 A The functional form of the potential is V 5 Er ri Pio y co 0 Oio bonds angles 2 2 ere EXimp xi gt Xi o T hey EXplanar Xi T Xi o So emsFol J escFol Oe backbone sidechains iy 12 iz 6 ono ol Y ane r r Tij contacts non contacts 38 Appendix A Energetic formulation 39 where Fp 1 cos di di0 5 1 cos 3 d d0 4 2 When using SMOG v2 all values may be adjusted by the user such as defining stabilizing non native interactions and including non specific dihedral angles However for the default model rio 9i 0 Xi o Gi and oj are given the values defined by the provided structure For the default model the parameters are set to the following values r 50 0 9 4060 Eximp 10 0 Expianar 40 0 ENC 0 160 ONC 2 5A
10. r Bjelkmar Rossen Apostolov Michael R Shirts Jeremy C Smith Peter M Kasson David van der Spoel Berk Hess and Erik Lindahl Gromacs 4 5 a high throughput and highly parallel open source molecular simulation toolkit Bioinformatics 29 7 845 854 2013 8 James C Phillips Rosemary Braun Wei Wang James Gumbart Emad Tajkhor shid Elizabeth Villa Christophe Chipot Robert D Skeel Laxmikant Kal and 47 Bibliography 48 9 10 11 Klaus Schulten Scalable molecular dynamics with namd J Comput Chem 26 16 1781 1802 2005 Jeffrey K Noel Paul C Whitford Karissa Y Sanbonmatsu and Jos N Onuchic SmogOctbp simplified deployment of structure based models in gromacs Nucleic Acids Res 38 W657 61 2010 Ohad Givaty and Y Levy Protein sliding along dna dynamics and structural characterization J Mol Biol 385 4 1087 97 Jan 2009 doi 10 1016 j jmb 2008 11 016 Jeffrey K Noel Paul C Whitford and Jos N Onuchic The shadow map a general contact definition for capturing the dynamics of biomolecular folding and function J Phys Chem B 116 29 8692 8702 2012 Heiko Lammert Alexander Schug and Jos N Onuchic Robustness and general ization of structure based models for protein folding and function Proteins 77 4 881 891 2009
11. default chain chainsFile Appendix B SCM jar 46 e Shadow map atomic contacts shadowing radius 2 A and cutoff 4 A user java jar SCM jar g protein gro t protein top o contactsDut m shadow c 4 s 2 chain chainsFile e Cutoff map atomic contacts and cutoff 4 A user java jar SCM jar g protein gro t protein top o contactsOut m cutoff c 4 chain chainsFile OR user java jar SCM jar g protein gro t protein top o contactsOut s shadow s 0 c 4 chain chainsFile e Shadow map residue contacts default include contact distances user java jar SCM jar g protein gro t protein top o contactsOut distances coarse CA chain chainsFile To calculate over a trajectory instead of a single structure use multiple X where X is the number of frames in the trajectory gro file Assumes that the format of proteinTraj gro is the same as the output of trjconv This saves time relative to looping over many grofiles because the topology and therefore the bonded list is only created once user GROMACS trjconv f traj xtc o proteinTraj gro user java jar SCM jar g proteinTraj gro t protein top o contactMapsOut default multiple 1000 chain chainsFile B 1 4 2 Full configuration parameter list The following will give a full list of configuration options user java jar SCM jar help B 1 4 3 Running SCM jar through the webtool On the webserver http smog server org
12. e g vi or emacs to prevent insertion of hidden characters e Only include lines that start with ATOM HETATM COMMENT may be at the beginning or end of any chain BOND user defined specific bonds Must appear after END TER to indicate a break between 2 chains and END Only BOND and COMMENT lines may appear after END e Chain identifiers are ignored If your system has multiple chains insert TER lines left justified between chains NOTE Do not immediately follow a TER line with an END line This is interpreted as a chain with 0 atoms and an error message will be issued Chapter 3 Usage 7 e Only residues and atoms within a residue defined in the forcefield templates will be recognized by SMOG v2 Unless a coarse grained template is designated with tCG unrecognized residues and atoms will lead to a PDB parse error and the program will exit 3 1 2 Preprocessing As discussed in Chapter 4 SMOG v2 reads template files in order to generate forcefield files As such each PDB file has to fully conform to the molecular structure definitions provided by the templates For example the default all atom templates provided in SBM_AA distinguish between terminal and non terminal residues i e in proteins there is an OXT in place of a peptide bond for terminal residues In the default templates the terminal amino acid residues have an suffix T added to their their three letter code e g GLY vs GLYT A prepr
13. server org Direct questions tot infolsmog server org SERCH EE dd EE dd dd dd dd dd dd EE EE EE EE EE EE dod In addition to verifying that SMOG v2 will start it is highly recommended that you also run the testing scripts provided as a tarball smog check tar which is available at smog server org smog check tar contains two main test scripts One script is very fast whereas the second is very comprehensive and can be used to test new SMOG models that you may design When everything works well performing the the checks is as easy as issuing two commands Just make sure you run config bash before running the tests While in the directory smog check issue the command quick check For the comprehensive check may take up to 30 minutes to complete smog check If you find that either script reports failures please communicate that to the smog server org team so that we may help diagnose the problem Chapter 3 Using SMOG v2 This chapter describes the usage of SMOG v2 It is recommended that all users read this chapter before using the software 3 1 Preparing the input PDB file 3 1 1 PDB file format To prepare a SMOG model a structural model e g crystallographic NMR or cryo EM model must be provided as a PDB file in accordance with the PDB Content Guide page 187 To avoid I O issues please follow these additional guidelines when preparing your PDB file for use with SMOG v2 e Only use a text editor
14. to all available bTypes For the case of the All Atom model since all the bType is identical we can instead also defined the bond interaction as shown under the lt bond gt tag in line 8 of Listing 4 5 Please note that the program will assign the interaction that most closely matches the bTypes of a given 4 atom pair One needs to be careful not to declare interactions that conflict with one another For example if your system contains a bond between atoms of bType B_1 and B_2 and bond definitions are only given for bTypes B_1 and B_2 then SMOG would not know which function to apply and it will exit with an error However if one also explicitly defined a bond function for B_1 B_2 pairs that bond would take priority The angle interaction follow a similar form as bonds but instead of expecting two b Type attributes it requires three The bType attribute in this case is symmetric to the central bType When matching bond angles the angle definition that matches the most atoms Q Oa FW NY FH Chapter 4 Template Based Approach 24 identically will be used Again if an equal number of atoms match in multiple angle definitions there would be ambiguity and SMOG will quit The dihedral_cosine function is classified under the dihedral directive with function type 1 in the sif file Listing 4 3 line 3 The arguments are of the form do a mult As introduced earlier the special character tells the program to ca
15. where all intra and inter molecular interactions as assigned minima that correspond to that structure This fully native centric variant of the model is colloquially referred to as a vanilla structure based model In terms of the energy landscapes of biomolecules these vanilla models represent an energetically unfrustrated landscape 3 4 Since biomolecular landscapes possess some degree of energetic rough ness it is often desirable to extend structure based models to include both native and non native interactions As such in the SMOG v2 software package we provide two en ergetically unfrustrated models by default upon which additional interactions may be added by the user Specifically in this software package we provide the coarse grained 1 1 available at http smog server org noel book_chapter sbm pdf 1 Chapter 1 Introduction 2 Co structure based model for proteins as developed by Clementi et al 5 We also provide the all atom structure based model as developed by Whitford et al 6 While the Ca model is only defined for proteins the all atom model supports proteins RNA DNA and some ligands While a complete description of the Ca model is available in the original reference there have been a number of extensions in the all atom model over the last several years Accordingly a complete description of the energetic parameters are given in Appendix A 1 2 What does SMOG v2 do SMOG v2 is a software p
16. 1 lt atom bType B_1 lt atom bType B_1 lt atoms gt lt bonds gt lt bonds gt lt impropers gt lt impropers gt lt residue gt lt RNA A gt nbT nbT nbT Type NB_1 Type NB_1 Type NB_1 nbType NB_1 nbT nbT Type NB_1 Type NB_1 nbType NB_1 nbT nbT nbT nbT nbT nbT nbT Pype NB_1 Type NB_1 Pype NB_1 Pype NB_1 Type NB_1 Type NB_1 Type NB_1 nbType NB_1 nbT Type NB_1 nbType NB_1 nbType NB_1 nbT nbT nbT nbT nbT nbT nbT Type NB_1 Type NB_1 Pype NB_1 Type NB_1 Type NB_2 Type NB_i Type NB_i nbType NB_1 nbT Pype NB_1 nbType NB_1 nbType NB_1 pairType P_1 gt P lt atom gt pairType P_1 gt 01P lt atom gt pairType P_1 gt 02P lt atom gt pairType P_1 gt 05 lt atom gt pairType P_1 gt C5 lt atom gt pairType P_1 gt C4 lt atom gt pairType P_1 gt 04 lt atom gt pairType P_1 gt C3 lt atom gt pairType P_1 gt 03x lt atom gt pairType P_1 gt C2 lt atom gt pairType P_1 gt 02 lt atom gt pairType P_1 gt C1 lt atom gt pairType P_1 gt N9 lt atom gt pairType P_1 gt C8 lt atom gt pairType P_1 gt N7 lt atom gt pairType P_1 gt C5 lt atom gt pairType P_1 gt C6 lt atom gt pairType P_1 gt N6 lt atom gt pairType P_1 gt N1 lt atom gt pairType P_1 gt C2 lt atom gt pairType P_1 gt N3 lt atom gt pairType P_1 gt C4 lt atom gt
17. 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 Chapter 5 Adding a new residue definition 32 lt ADDITIONAL TO lt bonds gt lt bond energyGroup pr_n gt lt atom gt N9 lt atom gt lt atom gt C4 lt atom gt lt bond gt FUNCTIONAL GROUP gt lt bond energyGroup pr_n gt lt atom gt C2 lt atom gt lt atom gt S10 lt atom gt lt bond gt lt bond energyGroup pr_n gt lt atom gt S10 lt atom gt lt atom gt C11 lt atom gt lt bond gt lt bond energyGroup pr_n gt lt atom gt N6 lt atom gt lt atom gt C12 lt atom gt lt bond gt lt bond energyGroup pr_n gt lt atom gt C12 lt atom gt lt atom gt C13 lt atom gt lt bond gt lt bond energyGroup pr_n gt lt atom gt C13 lt atom gt lt atom gt C14 lt atom gt lt bond gt lt bond energyGroup pr_n gt lt atom gt C14 lt atom gt lt atom gt C15 lt atom gt lt bond gt lt bond energyGroup pr_n gt lt atom gt C14 lt atom gt lt atom gt C16 lt atom gt lt bond gt LISTING 5 3 Adding the bonds section to the residue structure Note that the bonds are separated by the comments BACKBONE FUNCTIONAL GROUP ADDITIONAL BONDS TO RNA A FUNCTIONAL GROUP The bond tag attribute energyGroup classifies the energy group the specific bond belongs to and helps to determine the di hedral
18. AA model though there are a few minor changes First when running grompp you will want to change a few settings in the mdp file A sample mdp file for C models is given below integrator sd Run control Use Langevin Dynamics protocols dt 0 0005 time step in reduced units nsteps 100000 number of integration steps nstxout 100000 frequency to write coordinates to output trajectory trr file nstvout 100000 frequency to write velocities to output trajectory trr file nstlog 1000 frequency to write energies to log file nstenergy 1000 frequency to write energies to energy file nstxtcout 1000 frequency to write coordinates to xtc trajectory xtc_grps system group s to write to xtc trajectory assuming no ndx file is supplied to grompp energygrps system group s to write to energy file nstlist 20 Frequency to update the neighbor list coulombtype Cut off ns_type grid use grid based neighbor searching Chapter 3 Usage 13 rlist 3 0 cut off distance for the short range neighbor list rcoulomb 3 0 cut off distance for coulomb interactions rvdw 3 0 cut off distance for Vdw interactions coulombtype User vdwtype User pbc no Periodic boundary conditions in all the directions table extension 10 nm Should equals half of the box s longest diagonal tc grps system Temperature coupling tau_t 1 0 Temperature coupling time constant Smaller values stronger coup
19. CKBONE gt lt bond energyGroup bb_n gt lt atom gt P lt atom gt lt atom gt 01P lt atom gt lt bond gt lt bond energyGroup bb_n gt lt atom gt P lt atom gt lt atom gt 02P lt atom gt lt bond gt lt bond energyGroup bb_n gt lt atom gt P lt atom gt lt atom gt 05 lt atom gt lt bond gt lt bond energyGroup bb_n gt lt atom gt 05 lt atom gt lt atom gt C5 lt atom gt lt bond gt lt bond energyGroup bb_n gt lt atom gt C5 lt atom gt lt atom gt C4 lt atom gt lt bond gt lt bond energyGroup bb_n gt lt atom gt C4 lt atom gt lt atom gt 04 lt atom gt lt bond gt lt bond energyGroup bb_n gt lt atom gt C4 lt atom gt lt atom gt C3 lt atom gt lt bond gt lt bond energyGroup bb_n gt lt atom gt C3 lt atom gt lt atom gt 03 lt atom gt lt bond gt lt bond energyGroup bb_n gt lt atom gt C3 lt atom gt lt atom gt C2 lt atom gt lt bond gt lt bond energyGroup bb_n gt lt atom gt C2 lt atom gt 3020 3021 3022 3023 3024 3025 3026 3027 3028 3029 3030 3031 3032 3033 3034 3035 3036 3037 3038 3039 3040 3041 3042 3043 3044 3045 3046 3047 3048 3049 3050 3051 3052 3053 3054 3055 3056 3057 3058 3059 3060 3061 3062 3063 3064 3065 3066 3067 3068 3069 3070 3071 3072 3073 3074 Chapter 5 Adding a new residue definition 31 lt bond gt lt atom gt 02 lt atom gt
20. Gaussian potentials in the Go potentials section More information can be found in the NAMD manual A 3 4 Including Gaussian potentials in the topology files A 3 4 1 Gromacs The Gaussian interaction is designated in the pairs section of the topology file e ftype 6 Appendix A Energetic formulation 42 Cy A 1 4545 1009 St 1 ij c gt depth of the attractive well ro gt position of the minimum of the attractive well O gt width of the attractive well ryc gt position of the excluded volume hard wall This form includes an excluded volume part and therefore the pair ij should be included in exclusions The multiplicative form anchors the mini mum of the well at ro ec Note that ftype 5 and ftype 7 exist in the SBM extensions version though there is no implementation in SMOG2 at the moment They can be added by hand if desired A 3 4 2 NAMD Currently the nightly build version of NAMD contains the Gaussian potentials in the Go potentials section More information can be found in the NAMD manual Appendix B Understanding the provided SCM jar tool B 1 Introduction This section describes a Java application SCM jar that computes the Shadow map a general contact definition for capturing the dynamics of biomolecular folding and function It is described in the literature here 11 A contact map is a binary symmetric matrix that en
21. SBM_calpha gaussian A 3 1 templates SBM_AA gaussian Changes the contact potential to ftype 6 ec ro a u 50 In 2 ryc same as normal excluded volume ad defined by c12 such that i e onc ao The rather complex definition of the width of the Gaussian well o is designed to model the vari able width of the LJ potential Caa 1 2r r 1 2 so is defined such that G 1 2ri ri 1 2 giving o r 501n 2 Appendix A Energetic formulation Al A 3 2 templates SBM_calpha gaussian Changes the contact potential to ftype 6 ec ro 0 0 05 nm ruc 0 4 LJ6 12 LJ 10 12 Gaussian Multi basin FIGURE A 1 Comparison of Lennard Jones and Gaussian contact potentials Black curves show LJ contact potentials with minima at 6 A and 10 A The Gaussian contact potential shown in green has an excluded volume onc that can be set independently of the location of the minimum The dotted green line shows how the Gaussian contact would change as another minimum at 10 A is added Taken from 1 A 3 3 Downloading the source code extensions A 3 3 1 Gromacs The Gaussian contact shapes are not available in the standard Gromacs distributions The necessary source code can be obtained at http smog server org extension html This source distribution is compiled exactly as any other Gromacs source distribution A 3 3 2 NAMD Currently the nightly build version of NAMD contains the
22. SMOG Version 2 0 User s Manual This manual will continually be updated Last update July 20 2015 Rice University e Northeastern University Biomolecular Simulation Tool Authors Jeffrey Noel Mariana Levi Mohit Raghunathan Heiko Lammert Ryan Hayes Jos Onuchic Paul Whitford info smog server org Should you read this manual If you would only like to use the basic functionality of SMOG v2 i e the standard sup ported models then you may find that the README file associated with the distribution provides all the information you need This manual provides a more detailed description of the basic usage guidelines in addition to advanced usage information and detailed descriptions of the underlying methodologies models For basic users if the README is not sufficient then Chapters 1 2 and 3 will help you get started For more advanced users who may wish to modify structure based models e g extending to new residue types ligands electrostatics etc then consulting Chapters 4 and 5 will be necessary We additionally provide appendices that have technical details that may be of interest to some users While we try to provide all pertinent information here don t hesitate to contact us for clarification SMOG v2 and all associated files are distributed free of charge made available under the GNU General Public License Contents 5 1 1 Introduction 1 1 What are Structure Based Models o o
23. Shadow html one can build a shadow map from a SMOG formatted PDB file Bibliography 1 Jeffrey K Noel and Jos N Onuchic The many faces of structure based poten tials From protein folding landscapes to structural characterization of complex biomolecules Computational Modeling of Biological Systems Springer US pages 31 54 2012 2 Paul C Whitford Karissa Y Sanbonmatsu and Jos N Onuchic Biomolecular dynamics order disorder transitions and energy landscapes Rep Prog Phys 75 7 076601 2012 3 J Bryngelson and P Wolynes Intermediates and barrier crossing in a random energy model with applications to protein folding J Phys Chem 93 6902 6915 1989 4 J D Bryngelson J N Onuchic N D Socci and P G Wolynes Funnels pathways and the energy landscape of protein folding a synthesis Proteins 21 3 167 195 1995 5 C Clementi H Nymeyer and J N Onuchic Topological and energetic factors what determines the structural details of the transition state ensemble and en route intermediates for protein folding an investigation for small globular proteins J Mol Biol 298 5 937 953 2000 6 Paul C Whitford Jeffrey K Noel Shachi Gosavi Alexander Schug Kevin Y San bonmatsu and Jos N Onuchic An all atom structure based potential for proteins bridging minimal models with all atom empirical forcefields Proteins 75 2 430 441 2009 7 Sander Pronk Szil rd Pall Roland Schulz Per Larsson P
24. a non bonded group in the nb file Adding a new atom of a different mass requires a creation of a new non bonded group for the excluded volume interactions In our example we chose the atom sulfur to have larger mass of twice the carbon mass A new lt nonbond gt tag is added and it encapsulates the new non bond group information such as mass charge and other explicit non bonded terms The mass is doubled in the mass entry The nbType includes the previously defined group name NB_2 that is consistent with the bif file The modified file can be found in XMLcode_examples AA whitford09 withMIA nb Adding a new non bonded group in the nb fie hop O O ON DOF WY FH Bee FP eB No om FP W bb Chapter 5 Adding a new residue definition 35 lt xml version 1 0 gt lt nb gt lt DEFAULTS gt lt defaults gen pairs 0 gt lt GENERAL NONBONDS gt lt nonbond mass 2 00 charge 0 000 ptype A c6 0 0 c12 5 96046e 9 gt lt nbType gt NB_2 lt nbType gt lt nonbond gt lt nonbond mass 1 00 charge 0 000 ptype A c6 0 0 c12 5 96046e 9 gt lt nbType gt NB_1 lt nbType gt lt nonbond gt lt CONTACTS gt lt contact func contact_1 6 12 7 contactGroup c gt lt pairType gt lt pairType gt lt pairType gt lt pairType gt lt contact gt lt nb gt LISTING 5 5 Defining a new non bonded group in the nb file Chapter 6 Additional supported options As describe
25. ackage designed to allow the user to start with a structure of a biomolecule i e a PDB file and construct a structure based model which is then simulated using Gromacs 7 or NAMD 8 We previously implemented an online server http smog server org 9 that was capable of providing the vanilla flavor of structure based models along with a few adjustable parameters i e SMOG v1 SMOG v2 isa complete rewrite of the original software package and it provides four major advantages over its predecessor e Extensibility One may add new residue and molecule types without source code modifications e Portability By building forcefield definitions on generally defined XML formatted files researchers may easily distribute and share new SMOG model variants e Generalizability Every energetic parameter may be varied and additional ener getic interactions even non native may be included e Multi resolution capabilities All structural resolutions may be implemented as well as multi resolution variations It is important to note that none of these new features require additional programming or source code extensions Rather one simply adjusts the XML template files when designing SMOG variants Further the templates are not statically linked to SMOG v2 allowing the user to easily choose from a library of models at runtime Chapter 2 Installation Since SMOG v2 is written in Perl and Java there is n
26. be consistent with atom numbers in the input PDB file The fifth column can contain a numeric distance in A which if provided will be used instead of the native distance If using tCG to obtain a coarse grained topology the input contact map should designate residue numbers instead of atom numbers again with the same numbering in the input PDB file Chapter 3 Usage 11 3 4 Performing a simulation in Gromacs NAMD Once you have generate the top and gro files with SMOG you are ready to perform a simulation Rather than write a new molecular dynamics simulation package SMOG generates input files for use with Gromacs 7 and NAMD 8 two highly optimized and parallelized MD software suites This allows you to use nearly every protocol that has been implemented in these programs when performing simulations with structure based models e g replica exchange umbrella sampling In addition both of these packages are scalable to many processors through a combination of MPI and thread based parallelization allowing SMOG models to fully take advantage of cutting edge computing resources Here we provide brief descriptions of how to perform SMOG model simulations in Gromacs and NAMD More complete resources on performing these types of simulations may be found at http smog server org in the NAMD manual 3 4 1 Gromacs 4 5 or 4 6 3 4 1 1 All Atom Model First produce a portable xdr file in this case run tpr that describes your simulati
27. bonds as defined above highlighted in orange Finally we add all of the improper dihedrals to our residue In our example there is only one additional improper dihedral described above Add the improper dihedrals section to the residue structure lt impropers gt Chapter 5 Adding a new residue definition 34 3110 lt improper gt 3111 lt atom gt C3 lt atom gt 3112 lt atom gt C4 lt atom gt 3113 lt atom gt C5 lt atom gt 3114 lt atom gt 04 lt atom gt 3115 lt improper gt 3116 lt improper gt 3117 lt atom gt 03 lt atom gt 3118 lt atom gt C3 lt atom gt 3119 lt atom gt C4 lt atom gt 3120 lt atom gt C2 lt atom gt 3121 lt improper gt 3122 lt improper gt 3123 lt atom gt 02 lt atom gt 3124 lt atom gt C2 lt atom gt 3125 lt atom gt C1 lt atom gt 3126 lt atom gt C3 lt atom gt 3127 lt improper gt 3128 lt improper gt 3129 lt atom gt C2 lt atom gt 3130 lt atom gt C1 lt atom gt 3131 lt atom gt 04 lt atom gt 3132 lt atom gt N9 lt atom gt 3133 lt improper gt 3134 lt ADDITIONAL IMPROPER DIHEDRAL TO RNA A gt 3135 lt improper gt 3136 lt atom gt C13 lt atom gt 3137 lt atom gt C14 lt atom gt 3138 lt atom gt C15 lt atom gt 3139 lt atom gt C16 lt atom gt 3140 lt improper gt 3141 lt impropers gt 3142 lt residue gt LISTING 5 4 Adding the improper dihedrals section to the residue structure 5 4 Step 4 Define
28. codes which atom pairs are given attractive interactions in the SBM potential In the context of a SBM the native contact map should approximate the distribution of stabilizing enthalpy in the native state that is provided by short range interactions like van der Waals forces hydrogen bonding and salt bridges Any long range interactions or nonlocal effects are taken into account in a mean field way through the native bias B 1 1 Role of SCM jar in SMOG v2 Internally SMOG v2 uses SCM jar to compute contact maps From the user s point of view the contact map can be of two types all atom or coarse grained An all atom map returns the atoms that are in contact based on the Shadow definition A coarse grained map e g to be used with the Calpha model is created from an all atom map The coarse grained map consists of residue level contacts A residue level contact exists if there is at least one atom atom contact between two residues This is why a PDB containing all heavy atoms is required by the tool When coarse graining SMOG2 asks that the user provide an all atom template in addition to the coarse graining template that tells SMOG2 how to interpret the all atom PDB in order to interface with SCM jar 43 Appendix B SCM jar 44 FIGURE B 1 The Shadow contact map screening geometry Only atoms within the cutoff distance C are considered Atoms 1 and 2 are in contact because they are within C and have no intervening atom To chec
29. d above it is the aim of SMOG2 that the user will be able to extend the models in a wide range of ways Accordingly it is not possible that we provide descrip tions of every possible variation that one may explore However here we make an effort to provide some examples of how to implement specific features that we think may be frequently of interest 6 1 Adding specific bonds There are often cases where the covalent geometry of the system can not be determined by a general set of rules For example disulfide bonds may be formed or broken depend ing on the oxidation state As another example sugar structures often have branching patterns and therefore do not form linear chains For these types of chemical bonds we provide the BOND option in the PDB file If you would like to add a chemical bond then add BOND lines immediately after the END line in the PDB file The formatting is the following BOND ChainIndex1 AtomIndex1 ChainIndex2 AtomIndex2 energygroup ChainIndex1 and AtomIndex 1 indicate the first atom involved in the bond ChainIndex1 is the index of the chain in which the atom exists starting at 1 AtomIndex1 is the number of the atom as it appears in the PDB file Note ChainIndex1 is not necessarily the chain ID ChainIndex2 and AtomIndex2 indicate the second atom involved in the bond energygroup indicates the properties of any dihedral angles that have the new bond as a middle bond See Chapter 4 for discussion on energy g
30. ds gt lt function name bond_type6 directive bonds gt lt function name angle_harmonic directive angles gt lt function name angle_free directive angles gt lt function name dihedral_cosine directive dihedrals gt lt function name dihedral_harmonic directive dihedrals gt lt function name dihedral_free directive dihedrals gt lt function name contact_1 directive pairs exclusions 1 gt lt function name contact_2 directive pairs exclusions 1 gt lt function name contact_gaussian directive pairs exclusions 1 gt lt functions gt LISTING 4 3 Example of functions section of a sif file Each function is defined using the lt functions gt tag These function names are mapped to specific subroutines in src smogv2 To add a new interaction type follow the exam ples already in the code Note These functions are simply mappings to already defined Gromacs interactions If it doesn t exist in Gromacs it won t help to add it here The interactions currently in the code are listed in Table 4 2 If you add useful interactions please let us know so that we can incorporate them into the default distribution Group Settings The structure based model has two classes of energy groups contact groups and dihe dral groups Each contact group represents a collection of contacts and each dihedral Can be designated as a contact potential in nb for applications like restraining bound i
31. due Finally the optional attribute atomCount allows the user to explicitly set the total number of atoms to be counted for normalizing energies That is the total atom count is used in the energetic scaling procedure of dihedrals and contact energies as described in Appendix A This feature is useful when including many copies of a ligand in your system since the energetic normalization should only be based on the protein or RNA and not the multiply copied ligands In such a scenario the user would set atomCount to 0 If atomCount is not defined SMOG v2 automatically uses the total number of atoms listed under the lt atoms gt tag The lt atoms gt tag declares all the atoms in the residue Note that all the atoms within a specific residue in your PDB must be listed here If the PDB and bif are not consistent the program will terminate with an error This is the reason that the default templates differentiate between the C terminal and non terminal protein residues See Chapter 3 Each atom has a bond type bType a non bond type nbType and a pair type pairType The bond type attribute is used in the generation of the bonded interactions bonds angles and dihedrals Likewise the non bond type attribute is used in the generation of the non bonded interactions The pairType attribute is used in the generation of contact interactions 6 12 10 12 or Gaussian interactions OMAN OA FB wWN a an oF WN FO Chapter 4 Template Based App
32. e script will add one for you After you have generated your tabulated potentials for the 10 12 interaction i e ta ble file xvg and you have prepared a tpr file with grompp you can run the simulation with the command gt mdrun s run tpr noddcheck table table file xvg tablep table file xvg Typically for protein folding you will want to avoid domain decomposition and instead use particle decomposition by adding the option pd when on a single node After you perform your simulation you can utilize any analysis tools provided with Gromacs Chapter 3 Usage 14 3 4 1 3 Examples Check SMOG_PATH examples gromacs4 for some complete examples with terminal history 3 4 2 Notes and Hints 3 4 2 1 Domain Decomposition The pairs section is treated as bonded by Gromacs and therefore all pairs within a single domain are always calculated If you are using pd with version 4 X or only OpenMP threads in version 5 0 you can reduce the cutoffs to values that ignore the pair distances and only take into account the non bonded excluded volume provided you have no electrostatics of course For the default models this would be 0 6 for AA and 1 0 for CA 3 4 3 Gromacs 5 Gromacs 5 has a few changes that impact SMOG models First we don t yet provide a Gromacs 5 distribution with the SMOG enhancements umbrellas g kuh gaussian contact potentials So if you want to use these you can only use Gromacs 4 5 Gromacs 5 i
33. e name ALA residueType amino atomCount 5 gt lt atoms gt lt atom bType B_1 nbType NB_1 pairType P_1 gt N lt atom gt lt atom bType B_1 nbType NB_1 pairType P_1 gt CA lt atom gt lt atom bType B_1 nbType NB_1 pairType P_1 gt C lt atom gt lt atom bType B_1 nbType NB_1 pairType P_1 gt 0 lt atom gt lt atom bType B_1 nbType NB_1 pairType P_1 gt CB lt atom gt lt atoms gt lt bonds gt lt BACKBONE gt lt bond energyGroup bb_a gt lt atom gt N lt atom gt lt atom gt CA lt atom gt lt bond gt lt bond energyGroup bb_a gt 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 Chapter 4 Template Based Approach 18 lt atom gt CA lt atom gt lt atom gt C lt atom gt lt bond gt lt bond energyGroup bb_a gt lt atom gt C lt atom gt lt atom gt 0 lt atom gt lt bond gt lt FUNCTIONAL GROUP gt lt bond energyGroup sc_a gt lt atom gt CA lt atom gt lt atom gt CB lt atom gt lt bond gt lt bonds gt lt impropers gt lt improper gt lt atom gt CB lt atom gt lt atom gt CA lt atom gt lt atom gt C lt atom gt lt atom gt N lt atom gt lt improper gt lt impropers gt lt residue gt LISTING 4 1 Residue section of bif file The attribute name is the name of the residue as used in your PDB file The attribute residue residueType is the type of residue in this case an amino resi
34. ed Gromacs input files Important The all heavy atom geometry must be used even if the output will be a coarse grained residue based map for a Cg model The atomic co ordinates are read in gro format and the bond connectivity is read via a top obtained from the SMOG webtool or source distribution The topology is required since bonded atoms shadow each other differently and since contacts are automatically discarded be tween two atoms if they share a bonded interaction At the command line the basic syntax is user java Xmx1000m jar SCM jar g grofile t topfile o outputName chain chainFile default m shadow cutoff Xmx1000m assigns 1000 MB of RAM to the Java virtual machine heap With large complexes gt 1e5 atoms the default heap allocation can run out which gives the following error java lang OutOfMemoryError Java heap space The output all atom contact file format is chain_i atom_i chain_j atom_j distance and similarly the output residue contact file format is chain_i residue_i chain_j residue_j distance B 1 4 1 Some examples e Shadow map atomic contacts shadowing radius 1 A and cutoff 6 A default sizes See Figure B 1 for definition of radius and cutoff Add chain if you have multiple chains since the gro format does not allow for chain information Specify the chains file you get from your SMOG output user java jar SCM jar g protein gro t protein top o contactsOut
35. ed environment variables for your current session To automatically configure SMOG at login you may want to add the above command to your shell profile file e g bashrc source full smog2 path configure smog2 and add full smog2 path configure smog2 bin to your PATH 3 Verify that java is installed example output below gt java version java version 1 7 0_25 Java TM SE Runtime Environment build 1 7 0_25 b15 Java HotSpot TM 64 Bit Server VM build 23 25 b01 mixed mode If java is not found make sure to install the JRE or JDK 1 7 or greater and that it is in your path accessible as gt java 2 3 Verify SMOG is properly configured If SMOG is properly configured then you will be able to run smog with the following command Chapter 2 Installation 5 gt smog2 If configuration was successful then you will be greeted with the following message BRA AAR AAA AREA AEE EEE EE EE EE EE EE EE EE EE EE EE EE EE EE EE EE EE EE EE EE EE EE EE EEE EE EE EE EE EE EE EB a a ae dd dd deck k SMOG v2 0 sok dodo dedokokdciok dadokoidiciok k Thank you for using the Structure based Model SMOG software This package is the product of contributions from a number of people including Jeffrey Noel Mariana Levi Mohit Raghunathan Ryan Hayes Jose Onuchic amp Paul Whitford Copyright c 2015 The SMOG development team at Rice University and Northeastern University SMOG v2 0 amp Shadow are available at httpt smog
36. en by Appendix A Energetic formulation 40 Vca X xd Y erro gt gt co 0 00 Y enFol go bonds angles dihedrals cn AD Y as contacts non contacts where the dihedral potential Fp is 1 Fo 6 1 cos 6 1 cos 36 A 4 The coordinates X describe a configuration of the a carbons with the bond lengths to nearest neighbors r three body angles 0 four body dihedrals and distance between atoms 7 and j given by rij Protein contacts that are separated by less than 3 residues are neglected Excluded volume is maintained by a hard wall interaction giving the residues an apparent radius of onc 4 A The native bias is provided by using the parameters from the native state x9 Setting the energy scale e kp the coefficients are given the homogeneous values 100 ey 20e en c ENC A 3 Gaussian contact potential gaussian templates The Gaussian shaped contact potentials A 1 are available in the SMOG version of Gromacs and NAMD see section A 3 3 These potentials are used when one desires control over either the shape of the excluded volume or the width of the attractive potential They are also useful if contacts require minima in two places In depth characterization of the Gaussian potentials with all atom structure based models using SMOG can be found in 11 templates SBM_AA gaussian They are explored in the context of a multi basin Ca model here 12 templates
37. equired when using a coarse grained model The option tCG is used to indicate the precise coarse grained model that should be pre pared When tCG is given then the t flag is used to designate the templates that initially process the PDB for contact analysis Normally an all atom PDB is provided since native contact maps make the most sense when generated from an all atom struc ture note that the Shadow map only makes sense with atomic graining The tCG templates are then used to construct the CG energetic model In the above example the PDB has residues and atoms corresponding to the t templates and these definition will also be used for contact map generation See Appendix B for a detailed description of the supported contact map calculations 3 3 Input options SMOG v2 always requires a PDB file and some argument indicating that model should be used Table 3 1 shows the currently supported input arguments 3 3 1 User provided contact map If you have generated contacts yourself these can be used instead of using the internal SMOG2 routines A single file containing all the contacts in a list can be specified at the command line with the switch c For example gt smog2 i lt pdbfile gt c contacts txt will read the list of contacts in file contacts txt chainNum_i1 atomNum_i1 chainNum_j1 atomNum_j1 opt distance chainNum_i2 atomNum_i2 chainNum_j2 atomNum_j2 opt distance Chapter 3 Usage 10
38. h intraRelativeStrength 1 and normalize option set to 1 true Chapter 4 Template Based Approach 26 SMOG2 supports two types of interactions for native contacts Since the contact inter actions are unique to SMOG v2 refer to Table 4 2 for how SMOG2 expects the input parameters In code Listing 4 7 the contact parameters eq c6 and c12 are calculated automatically by the program using energy normalization and the PDB structure be cause of the 3 ec is also calculated dynamically through the scaling equations If one uses the 10 12 or N M functions table files have to be included see Section 3 4 1 2 for syntax and details Chapter 5 Adding a new residue This chapter provides a step by step tutorial on how to add a new residue type using SMOG v2 template files 5 1 Step 1 Examine the molecular structure The residue 2 methylthio N6 isopentenyl adenosine MIA is a modified nucleic acid residue that is present in many RNA structures It is identical to Adenine except there are a few additional carbon atoms and a sulfur atom ligated to XXX For the purposes of demonstrating We would also like to define a larger mass for sulfur atom Make sure to have the correct chemical structure of your molecule A useful method is to visually inspect it with a molecular visualization program VMD for example 27 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 Chapter 5 Adding a new re
39. hedrals the second input parameter for a type 1 dihedral is the energetic scaling term The option is used to tell the program that the dihedral is considered under the global energy normalization procedure g is dynamically calculated by the program A note on assigning dihedrals As with bonds and angles it is quite easy to provide multiple b Type sequences that will match to the same atoms in a system For example if you define a dihedral function for B_1 B_2 as well as B_1 then a dihedral in your system between atoms B_1 B_2 B_1 B_1 would match both To determine which OMAN DA AUNE hop o Chapter 4 Template Based Approach 25 function should be applied a scoring function S is used S is defined as 2 times the number of exactly matching b Types plus 1 times the number of wildcard matches If any one of the atoms does not have an exact match nor a wildcard then S 0 According to this criteria S 6 for the first definition and S 5 for the second Accordingly the B_1 B_2 definition would take priority If two dihedrals have equal S values S 0 for all dihedral functions then the program will quit with an error Non Bonded Interactions The nb files is used to generate non bonded interactions such as native contacts and non specific interactions lt DEFAULTS gt lt defaults gen pairs 0 gt lt GENERAL NONBONDS gt lt nonbond mass 1 00 charge 0 000 pty
40. k if atoms 1 and 3 are in contact one checks if atom 2 shadows atom 1 from atom 3 The three atoms are viewed in the plane and all atoms are given the same shadowing radius S Since a light shining from the center of atom 1 causes a shadow to be cast on atom 3 atoms 1 and 3 are not in contact The tool is available with the SMOG2 source for users that want to create their own customized maps The rest of the chapter describes the basics of using the tool B 1 1 1 Some details of coarse graining The coarse grained contact map returned is only strictly recommended for use with Calpha models of proteins and where the input PDB has an all atom representation For various modeling applications it is desirable that the program not die with an error if the PDB doesn t only contain all atom protein with each residue containing a CA atom Therefore the behavior is that the program will choose one atom from each residue to stand in as the representative coarse grained position It chooses in order of preference CA N1 first atom in the residue This really only matters for the distance option B 1 2 Locating SCM jar The jar should be located in MOG_PATH tools B 1 3 Citing SCM jar The citation for SCM jar is 11 Appendix B SCM jar 45 B 1 4 Running SCM jar Like any java application no compilation is necessary but a virtual machine is required SCM jar requires a sufficiently recent JRE SCM jar reads SMOG formatt
41. lculate the native value from the PDB structure file In this case we ask the program to calculate the dihedral angle for all dihedral interactions that involve the bType combination By using the argument along with a multiplicity factor third argument we tell the program to multiply all the dy values by the multiplicity factor More generally one may provide any function for the angle calculation For example if you were to use 2 1 the angle would be evaluated as native angle times 2 and minus 1 Fp 9 1 cos po 0 5 1 cos 3 o 4 6 Equation 4 6 shown above is the dihedral interaction function used in the All Atom model The code Listing 4 6 shown below shows how to define the second term of Fp lt dihedral func dihedral_cosine 3 dihedral_cosine 1 energyGroup bb_n gt lt bType gt lt bType gt lt bType gt lt bType gt lt bType gt lt bType gt lt bType gt lt bType gt lt dihedral gt LISTING 4 6 Dihedral section of b file Finally multiple function can be applied to the same dihedral angle by including a sum of functions e g func f g h The special keyword has limitations in where it can be used For bonds and angles it can only be used for the first input parameter to a function For dihedrals declared as type 1 it can be used for the first two input parameters to a function In the case of di
42. lecules to be supported Setting Information File sif Defines interaction function declarations Bond Information File b Defines bonded interactions between atoms Nonbond Information File nb Defines non bonded interac tions between atoms TABLE 4 1 Descriptions of the 4 files that comprise a single template The expected suffix of each file is shown in parentheses your template folder contains more than one file of a specific file type the program will exit with an error Each file contains unique information as described below 4 2 1 Biomolecular Information File bif The Biomolecular Information File from here on called bif defines the covalent struc ture of all residues described by a particular forcefield The bif file is used to extract the appropriate coordinate information from the PDB file Since the PDB file only provides the coordinates numbers and names of atoms and residues the residue definitions in the bif file are used in conjunction with the PDB file to build a forcefield for a particular biomolecule Each residue is defined in the bif file by declaring all the atoms in that residue the bonds between the atoms and the improper dihedrals between the atoms Residues Each residue is individually defined between the lt residues gt and lt residues gt tags As an example the text below shows how one would define the residue ALA which contains 5 atoms lt residu
43. ling ref_t 120 0 In reduced units see Gromacs manual for details Pcoupl no Pressure coupling gen_vel yes Velocity generation gen_temp 50 0 gen_seed 1 ld_seed 1 comm_mode angular center of mass velocity removal LISTING 3 2 Sample mdp file for Ca SMOG models used for Gromacs v4 5 The most significant difference is the use of User defined Vdw and Coulomb inter actions This is due to the fact that the 10 12 potential used for contact interactions in the Cy model In order to run mdrun next step it is necessary to generate table files that define the 10 12 interaction We provide a tools for generating these tables SMOG_PATH bin smog tablegen with the SMOG2 distribution The table can be generates in a single step SMOG_PATH bin smog tablegen lt M gt lt N gt lt ion conc gt lt elec switch dist gt lt elec truncate dist gt lt table length gt lt output name gt M the exponent on the attractive term N is the exponent on the repulsive term If you are not including electrostatics most common then provide values of 0 for lt ion conc gt lt elec switch dist gt and lt elec truncate dist gt lt table length gt indicates how long the table should be in nanometers It is important that the table is longer than any native contacting pair of atoms may be during the simulation Finally the last argument is the filename for the table file If you don t use a xvg suffice th
44. ls e g the 10 12 potential used for Ca models See Chapter 3 Usage SMOG_PATH bin smog tablegen lt M gt lt N gt lt ion conc gt lt elec switch dist gt lt elec truncate dist gt lt table length gt lt output name gt lt ion conc gt is the monovalent ion concentration which sets k In order to avoid numerical issues by imposing a cutoff distance lt elec switch dist gt and lt elec truncate dist gt should be used These will modify the potential starting at a dis tance lt elec switch dist gt using a 4th order polynomial which will ensure that the potential smoothly switches to zero at lt elec truncate dist gt Caution When us ing electrostatics you will need to adjust your mdp so that cutoffs are properly imposed during the simulation The above flags only ensure that the table is correct Appendix A Energetic Description of the Distributed Models A 1 The All atom model All non hydrogen atoms are explicitly represented and the provided structure file pdb is defined as the global potential energy minimum Here we provide a complete de scription of the default all atom structure based model energy function which is defined by the template SBM_AA All calculations employ reduced units Each atom is rep resented as a single bead of unit mass and the charge of each atom is set to zero Covalent geometry is maintained through harmonic interactions that ensure the bond lengths bond angles
45. lude non specific interactions The plug and play nature of the templates has the additional advantage of forcefield portability and easy sharing of user created variations of structure based potentials A single SMOG template is comprised of four XML formatted files These files are absolutely necessary when using SMOG v2 XML format was adopted because of its consistent formatting ease of editability and readability and there are widely available program modules to generate and parse XML files Furthermore XML allows for schemas a form content format restriction file to which the template files must conform which adds an additional layer of error checking capabilities This chapter assumes that the user knows the basics of XML formatting Users unfamiliar with XML formatting may want to check out the W8schools website Table 4 1 summarizes the purpose of each template file In Chapter 5 we show how to add new new residue to the template files for an All Atom structure based model 4 2 SMOG v2 Templates SMOG v2 expects four template files to be present in a single folder i e the template folder As discussed in Chapter 3 the template folder name is a required argument when running SMOG v2 A template folder can only contain one of each file type If 16 OMAN OA FE wWN Rh bh oF O Ne O Chapter 4 Template Based Approach 17 File Purpose Biomolecular Information File bif Defines the structure of biomo
46. o need for compilation and instal lation However one must configure a few settings and ensure that appropriate modules are available at runtime This section describes the steps necessary to configure and verify that SMOG v2 is functioning properly on your local machine 2 1 Prerequisites SMOG v2 runs on all Unix like operating systems The prerequisites for SMOG v2 are Perl Programming Language Perl Data Language PDL as well as the following modules which are available through the Perl module managing utility CPAN recommended or through manual installation String Util XML Simple Exporter XML Validator Schema Finally your machine must have Java Runtime Environment v1 7 or greater 2 2 Configuration Before running SMOG v2 you must configure it on your local machine This is accom plished through a short two step process Chapter 2 Installation 4 1 Set the required environmental variables To do so modify the file configure smog2 which is included with the distribution Specifically you will need to modify the follow ing two lines smog2dir perl4smog smog2dir is the main SMOG directory and perl4smog is the version of perl that you would like to use On most linux systems the default location of Perl is usr bin perl whereas on OSX it is typically opt local bin per1 2 Initialize the new environment variables with gt source configure smog2 This will set the requir
47. ocessing tool smog _adjustPDB is provided that will adjust your PDB to reflect changes necessary to conform to the templates Assuming SMOG_PATH bin is in your PATH the preprocessing tool may be run with the following command gt smog2_adjustPDB lt PDB file gt lt default f mapFile gt lt outputPDB pdb gt The first required input is the PDB file the second required input is either the op tion default to use the map file provided with the program for use with the default SMOG models or the option f followed by your own map file The map file should be formatted as follows mapFile lt Residue gt lt head terminal gt lt tail terminal gt ALA ALA ALAT G G5 G Lines containing a character are interpreted as comments Each line must have three strings that are space tab delimited The first field is the original residue name as it appears in the original PDB file The second is the name to be substituted if the residue is the first residue in a chain and the third field is the corresponding substitution for the last residue in each chain The preprocessing tool will write a modified PDB file Chapter 3 Usage 8 outputPDB pdb The script also numbers atom and residue indices to be sequential within each chain and it adjusts atom names to be consistent with the SBM_AA template files 3 2 Generating a Structure Based Model SMOG v2 supports a broad range of structure based models and the All Atom
48. ome examples B 1 4 2 Full configuration parameter list B 1 4 8 Running SCM jar through the webtool 28 28 28 29 30 33 34 36 36 37 38 38 39 40 40 41 41 41 41 41 41 42 43 43 43 44 44 44 45 45 46 46 47 Chapter 1 Introduction 1 1 What are Structure Based Models Structure based models i e SBMs or SMOG models define a particular known con formation as a potential energy minimum With this being the only requirement there is an endless number of ways in which one may construct a structure based model For example one may build protein specific and RNA specific variants the resolution of the model can be varied multiple minima may be included and the degree to which non native interactions are stabilizing can be adjusted The utility of these models is equally diverse where they may be applied for understanding dynamics or for structural modeling objectives as discussed in recent review articles 1 2 With such flexibility variations within this general class of models can be tailored to ask specific questions about biomolecular processes In the present document we describe a set of computa tional tools that allows one to use previously developed structure based models as well as design and implement new variations that are suited for your specific needs In the simplest form a structure based model defines a single configuration as the global potential energy minimum
49. on This file is platform independent and contains all parameters for your simulations This allows you to produce a tpr file on any machine and then move it to another machine and run your simulation The xdr file is produced by grompp part of the Gromacs distribution gt grompp f mdpfile mdp c gro_file gro p top_file top o run tpr The file mdpfile mdp tells Gromacs what settings to use during the simulation such as the timestep size the number of timesteps and what thermostat to use Here is a sample set of configurations that are consistent with the default all atom model integrator sd Run control Use Langevin Dynamics protocols dt 0 002 time step in reduced units nsteps 100000 number of integration steps nstxout 100000 frequency to write coordinates to output trajectory trr file nstvout 100000 frequency to write velocities to output trajectory trr file nstlog 1000 frequency to write energies to log file nstenergy 1000 frequency to write energies to energy file nstxtcout 1000 frequency to write coordinates to xtc trajectory xtc_grps system group s to write to xtc trajectory assuming no ndx file is supplied to grompp energygrps system group s to write to energy file nstlist 20 Frequency to update the neighbor list coulombtype Cut off Chapter 3 Usage 12 ns_type grid use grid based neighbor searching rlist 1 2 cut off distance for the short range neighbor list
50. ons or elastic network models 2Minimum at native distance if using for c6 c12 3Minimum at native distance if using for cN cM Need to include a table file with tablep with mdrun Details in Appendix A 3 4 1 Chapter 4 Template Based Approach 21 Name top ftype Input Parameters bond_harmonic bonds 1 ro k bond_type6 bonds 6 ro k angle_harmonic angles 1 Oo k dihedral_harmonic dihedrals 2 Qo k dihedral_cosine dihedrals 1 oo k multiplicity contact_1 pairs 1 N M c6 c12 contact_2 pairs 1 N M cN cM contact_gaussian pairs 6 EC TNC O TO TABLE 4 2 Functions available in SMOG2 group represents a collection of dihedrals These energy groups are used for energetic scaling of interaction strength Haa Dd espFo escFb 4 1 backbone sidechain y ecFeontacts Y ds 4 2 contacts Shown above is the All Atom Hamiltonian with only the dihedral and contact terms Shown below are the energetic scalings of the dihedral and contact terms and their respective attributes under the sif file eBB _ intraRelativeStrength bb 4 3 ESC intraRelativeStrength sc y EBB y Esc y ec Total non ligand atoms 4 4 gt eBB esc dihedrals groupRatio 4 5 Xec contacts groupRatio The program automatically calculates the total number of non ligand atoms used in the energetic scaling Although the scaling equation
51. pe A c6 0 0 c12 5 96046e 9 gt lt nbType gt NB_1 lt nbType gt lt nonbond gt lt CONTACTS gt lt contact func contact_1 6 12 7 contactGroup c gt lt pairType gt lt pairType gt lt pairType gt lt pairType gt lt contact gt LISTING 4 7 Contacts section of nb file Listing 4 7 shows how to define non bonded and native contact parameters The nb Type value defined for each atom in the bif is matched to the nonbond declaration Note the contacts are placed in the pairs section in the topology file which was originally used for 1 4 pair interactions For structure based models it has been applied to include native contacts For structure based models a non bonded interaction is a volume exclusion interaction usually defined as the Lennard Jones 12 term Gromacs generates these interactions using the information provided in the defaults and the atomtypes sections The non bonded interaction declaration in SMOG2 contains the atom attributes mass charge and atom type as well as the explicit c6 and c12 terms for the Lennard Jones function If one wanted to include non specific attractive interactions between atoms then a non zero value should be given for the c6 parameter A contact function declaration includes the additional contactGroup attribute which is used to map the function to specific groups of contacts For example in Listing 4 4 the contact group c was declared wit
52. propers gt lt residue gt lt RNA A gt lt residue name A residueType nucleic gt LISTING 5 1 Nucleic residue section of bif file Keep in mind that the attribute name should match the residue name in the PDB file 2945 2946 2947 2948 2949 2950 2951 2952 2953 2954 2955 2956 2957 2958 2959 2960 2961 2962 2963 2964 2965 2966 2967 2968 2969 2970 2971 2972 2973 2974 2975 2976 2977 2978 2979 2980 2981 2982 2983 2984 2985 2986 Chapter 5 Adding a new residue definition 29 5 3 2 List all of the atoms in the residue List all of the atom names in your residue as they appear in your PDB The lt atoms gt tag encapsulates all the atoms in the biomolecule lt NUCLEIC RESIDUES gt lt 2 methylthio N6 isopentenyl adenosine gt lt residue name MIA residueType nucleic gt lt atoms gt lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_1 lt atom bType B_
53. roach 19 The lt bonds gt tag contains all the bonds that should be present in the residue Each bond in a residue is listed under the lt bond gt tag The atom names must match those listed in the lt atoms gt field The bond tag also has an attribute called energyGroup that allows for one to define heterogeneous energetics in the system The energy group attribute is used in conjunction with the bond types to determine the dihedral interaction Using the bonds declared here the program dynamically calculates all angles and dihedrals that can exist in the molecule The lt impropers gt tag contains all the improper dihedral angles in the biomolecule The tag lt improper gt lt improper gt holds four atom tags The order of the four atoms defines a specific improper dihedral within a residue This feature is used to add dihedrals that cannot be determined based on bond geometry Connections In addition to defining a residue the bif file is also used to define how sequential residues are covalently connected Listing 4 2 1 shows how two residues of type amino are covalently linked The attribute residueTypel and residueType2 declares how a residue of type residueTypel at position n should be connected to a residue of type residueType2 at position n 1 The residue types are matched based on the residue definitions Much of the structure of the connection element is similar to that of the residue element There is a single bond whereb
54. roups For example the following line BOND 1 51 4 100 r_p 36 Chapter 6 Additional options 37 would add a chemical bond between the atom numbered 51 in the first chain and the atom numbered 100 in the 4th chain The bond properties would be determined based on the b file settings and the energy group of any associated dihedrals would be r_p When SMOG2 runs it will write information to the screen as the BOND lines are detected It will also write out information about what it interpreted the lines to mean so you can verify that the intended bonds are being added 6 2 Adding electrostatics and non standard contact poten tials While not part of the standard structure based model there are often times where it is desirable to add some degree of electrostatic interactions or one would like to use interactions that are not of the 6 12 form If you would like to add charge to your system then you will need to define atom types with varying charges which is accomplished by modifying the nb file See Chapter 4 By default Gromacs will treat these interactions as purely Coulombic If you would like to use a screened electrostatic interaction i e Debye Huckle then you need to supply a table We supply a tool that will generate a screened electrostatic look up table as implemented by Givaty and Levy 10 To generate this table you can use the tool smog_tablegen This is the same tool used for non standard N M potentia
55. s shown above is limited to residue types with only two dihedral types backbone and sidechain dihedrals and single contact type the program allows for scaling equations to be generalized to more than two dihedral types and more than one contact type The energy group ratios are contained within the lt Groups gt tag lt energyGroup name bb_n residueType nucleic intraRelativeStrength 1 normalize 1 gt NO of Wh Chapter 4 Template Based Approach 22 lt energyGroup name sc_n residueType nucleic intraRelativeStrength 1 normalize 1 gt lt energyGroup name pr_n residueType nucleic normalize 0 gt lt energyGroup name ip_n residueType nucleic normalize 0 gt lt energyGroup name r_n residueType nucleic normalize 0 gt lt contactGroup name c intraRelativeStrength 1 normalize 1 gt lt groupRatios contacts 2 dihedrals 1 gt LISTING 4 4 Energy group section of sif file The two classes of energy group ratios dihedral and contact ratios are controlled under the lt energyGroup gt and lt contactGroup gt tags respectively The residueType attribute is used to designate the residue type the scaling factors of a particular energy group is used for The name attribute is the label for the energy group The name attribute used here is matched to the energyGroup attribute under lt bond gt tag in the bif file The name of a particular contact energy group will be used la
56. sidue definition 28 5 2 Step 2 Create a new All Atom template directory Since we re going to explicitly define each atom in the MIA residue the model used here will be All Atom model You can modify the default All Atom directory provided by Smog2 SBM_AA or create a copy of it in your working directory to make the changes in the template files In this example we will modify the existing AA whitford09 bif and AA whitford09 nb files 5 3 Step 3 Define a new residue As mentioned in chapter 4 the biomolecular information file bif declares the struc ture of all biomolecules in your system Here we will define the residue information by declaring all of the atoms bonds and improper dihedrals within the residue The full modified file can be found in the XMLcode_examples AA whitford09 withMIA bif directory 5 3 1 Place the new residue tag in the bif file As you get acquainted with the AA whitford09 bif file structure you ll find that the residues appear grouped together according to their type Ligands amino and nucleic residues The residue type of MIA is nucleic so it is added for convenience next to the existing nucleic residues The lt residue gt tag encapsulates all of the residue information lt NUCLEIC RESIDUES gt lt 2 methylthio N6 isopentenyl adenosine gt lt residue name MIA residueType nucleic gt lt atoms gt lt atoms gt lt bonds gt lt bonds gt lt impropers gt lt im
57. strengths The energy group bb_n refers to bonds that belong to the backbone group and pr_n refers to the functional group It is sometimes useful to use an existing residue as a reference if we know that the new residue is a slight modification of it In our example MIA is a modified RNA A molecule All backbone and functional group bonds of RNA A can be added to MIA and we are left to determine the few other bonds created by the additional atoms S10 C11 C16 The additions are added under the comment ADDITIONAL BONDS TO RNA A FUNCTIONAL GROUP with energy group of pr n 3109 Chapter 5 Adding a new residue definition 33 5 3 4 List the improper dihedrals Improper dihedrals cannot be dynamically calculated by the program using the bonds and should be added separately The tag lt improper gt lt improper gt holds four atoms The order of the four atoms here defines a specific improper dihedral within a biomolecule How to find an improper dihedral An improper dihedral is used to ensure proper geometry about a chiral centers i e prevent symmetry inversion due to an absent hydrogen atom A proper dihedral would be defined by four sequential atoms connected by bonds such as the dihedral C4 C5 05 P An improper dihedral is defined by atoms that are not sequential Those angles need to be identified using the chemical structure of the molecule For example we consider the four carbon atoms C13 C14 C15 C16 and their
58. tep 1 Examine the molecular structure 0004 ii 16 16 16 17 20 22 27 iii 5 2 Step 2 Create a new All Atom template directory 5 3 Step 3 Define a new residue 5 3 1 5 3 2 5 3 3 5 3 4 Place the new residue tag in the bif file List all of the atoms in the residue List all of the atom bonds List the improper dihedrals 5 4 Step 4 Define a non bonded group in the nb file 6 Additional supported options 6 1 Adding specific bonds 6 2 Adding electrostatics and non standard contact potentials A Energetic Description of the Distributed Models A 1 The All atom model A2 The to Model cc gn oe wl pro a ld A 3 Gaussian contact potential gaussian templates A 3 1 A 3 2 A 3 3 A 3 4 templates SBM_AA gaussian templates SBM_calpha gaussian Downloading the source code extensions Aal IDAS uo oia o eh e Aldaz NAMD ses redta e Including Gaussian potentials in the topology files A 3 4 1 Gromacs A342 NAMD ociosas B Understanding the provided SCM jar tool B1 Introductio se 2 266 e a ee Bled B 1 2 B 1 3 B 1 4 Bibliography Role of SCM jarinSMOGv2 B 1 1 1 Some details of coarse graining Locating SCM Jar sc ra ee a Ok ee Citing SCN IRE e sas a gai a sika ee as Running SOM JUE ce e aoea a ie we i B 1 4 1 S
59. ter when declaring contact interaction functions in the subsequent section of this chapter The normalize attribute for each energy group is a boolean attribute 1 or 0 and is used to determine if a particular energy group is to be included in energy normalization see equation 4 4 For the All Atom model the dihedral group with the name pr_n which represents the nucleic planar rigid dihedrals has a normalize option set to 0 indicating that planar dihedral in nucleic acids will not be part of the normalization In contrast in the All Atom model sidechain dihedrals are normalized as are backbone dihedrals and contact energies Accordingly those energy groups have the normalize option set to 1 The intraRelativeStrength attribute is the relative ratio of stabilizing energy within the different class of energy group for a particular residue see equations 4 3 Finally we use the lt groupRatios gt tag to set the energy partition between the two classes of energy group according to equation 4 5 4 2 3 Bond Information File b amp Nonbond Information File nb The b file is used to define all bonded interactions which includes bonds angles and dihedral The nb file is used to define all the non bonded interactions such as contacts 1 4 pairs and excluded volume Bonded Interactions We first discuss how to define a basic bond interaction for the All atom model Chapter 4 Template Based Approach 23 lt
60. tself has changes of note 1 OpenMP support has replaced the option of particle decomposition and 2 OpenMP requires cutoff scheme Verlet and Verlet doesn t yet allow tabulated potentials This has the largest impact on C alpha models which use tabulated potentials If your simulated system has less than roughly 100 atoms you can typically only use a single processor with v5 because additional threads are only allowed through OpenMP If your system is large enough you can use multiple MPI processes with domain decomposition to scale to multiple cores When using Verlet lists you have to use pbc xyz For all atom simulations Verlet lists are fine and it is usually best to use as many OpenMP threads as possible with ntomp 3 4 3 1 Examples Check SSMOG_PATH examples gromacs5 for complete examples including terminal history Chapter 3 Usage 15 3 4 4 NAMD The forcefield files generated by SMOG v2 are fully compatible with NAMD To perform SMOG models in NAMD please consult the NAMD manual A tutorial is available http vidar scs uiuc edu jlai7 Tutorial GoDemo pdf Chapter 4 Template Based Approach 4 1 Introduction to templates SMOG v2 offers increased versatility over SMOG v1 9 by shifting to a template based approach for defining molecular structures Each template allows for direct control of the structure based energy function which may include but is not limited to multi resolution models and models that inc
61. y the first atom belongs to the nth residue and the second atom belongs to the n 1 th residue We can also define though not a requirement component of all connection definitions a single improper dihedral In the context of impropers the special character suffix is used to declare atoms that belong to the n 1 th residue In the code listing the N atom belongs to the n 1 th residue lt connections gt lt AMINO AMINO gt lt connection name amino amino residueTypel amino residueType2 amino gt lt bond energyGroup r_a gt lt atom gt C lt atom gt lt atom gt N lt atom gt lt bond gt lt improper gt lt atom gt 0 lt atom gt lt atom gt CA lt atom gt lt atom gt C lt atom gt lt atom gt N lt atom gt lt improper gt lt connection gt Oo ooN ADA FB wWN re e NF O Chapter 4 Template Based Approach 20 LISTING 4 2 Connection section of bif file 4 2 2 Setting Information File sif While the bif file is used to define the covalent geometry of each residue the Setting Information File sif is used to control the distribution and functional form of the energetics which includes the inter dihedral dihedral ratios contact to dihedral ratios contact map settings and function declarations Functions The lt functions gt tag should list all the functions that the model requires lt functions gt lt function name bond_harmonic directive bon

Download Pdf Manuals

Related Search

Related Contents

Home Decorators Collection 1890140610 Use and Care Manual  ワンタイムパスワードご利用マニュアル（ソフトウェアトークン編）  operating manual manual de instrucciones manuel d`utilisation  FEIN 72294264090 Use and Care Manual  Kenwood HM-437MP Stereo System User Manual  D-Link DCS-942L  GUIDA PER L`UTENTE - Esse-ti  クラスIII免疫・内分泌検査用シリーズ サイログロブリン自己    Fujitsu ESPRIMO P900  

Copyright © All rights reserved. Failed to retrieve file