confinedBSA.ipynb

{
 "metadata": {
  "name": "",
  "signature": "sha256:9c29956d28f34700ff4ef3b8b6a2269a97ad6c7738958bb2609a5acc7aa89718"
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "<b>Note:</b> Best viewed in the [notebook viewer](http://nbviewer.jupyter.org/github/jmborr/confinedBSA/blob/master/confinedBSA.ipynb)\n",
      "<a id='Table of Contents'></a><h2>Confined BSA</h2>  \n",
      "<a href='#goal'>Goal</a>   \n",
      "<a href='#people'>People</a>  \n",
      "\n",
      "<a id='Experiment'></a><h3>Experiment</h3>\n",
      "<a href='#Experiment.details'>Details of the experiment</a>  \n",
      "\n",
      "<a id='Simulations'></a><h3>Simulations</h3>\n",
      "<a href='#Simulations.references'>References</a>  \n",
      "<a href='#Simulations.gromacs_tutorial'>Gromacs tutorial for solvated Lyzosime</a>  \n",
      "<a href='#Simulations.solvatedMBSA'>Solvated Monomeric BSA box</a>  \n",
      "<a href='#Simulations.amorphous_from_MS'>Amorphous silica box from Materials Studio post</a>  \n",
      "<a href='#Simulations.cristobalite_2_2_2'>Simulations of small cristobalite crystal</a>  \n",
      "<a href='#Simulations.cristobalite'>Cristobalite</a>  \n",
      "<a href='#Simulations.confining_cristobalite'>Protein confined in cristobalite</a>  \n",
      "<a href='#Simulations.comparing_OA_ovl'>Comparing confinements and OA charges</a>  \n",
      "<a href='#Simulations.confining_cristobalite_q0.90ovl2.5'>Protein confined in cristobalite with $q_{OA}=0.90$ and overlap=2.5A</a>  \n",
      "<a href='#Simulations.interfaceFF'>The INTERFACE force field</a>  \n",
      "<a href='#Simulations.ZhangFF'>Zhang's force field</a>  \n",
      "\n",
      "<a href='#Simulations.protocol'>Protocol for initial configuration for confinedBSA</a>  \n",
      "\n",
      "<a id='meetings and reports'></a><b>Meetings</b>  \n",
      "<a href='#meetings.20160812'>Example meeting</a>  \n",
      "\n",
      "<a href='#Syntax'>HTML and Markdown Syntax Examples</a>  \n"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "(<a href='#Table of Contents'>Top</a>)<a id='goal'></a><h3>Goal</h3>\n",
      "Simulate Bovine Serum Albumin (BSA) confined in a silica nanopore and compare dynamics to results from QENS experiments. The goal of the experiments is to investigate the dynamics of single protein cluster under ultra-confinement using BASIS"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "(<a href='#Table of Contents'>Top</a>)<a id='people'></a><h3>People</h3>\n",
      "Eugene Mamontov  \n",
      "Kevin Slater  \n",
      "Panchao Yin  "
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "(<a href='#Experiment'>Top</a>)<a id='Experiment.details'></a><h3>Details of the experiment</h3>\n",
      "\n",
      "IPTS-18245 - [proposal document](files/experiment/review_sor18245.1.pdf)\n",
      "\n",
      "Two samples:  \n",
      "- DBSA: encapsulated deuterated water and hydrogenated protein.\n",
      "- HBSA: dry protein powder, containing no water (or just crystallographic) and hydrogenated protein.\n",
      "\n",
      "Susceptibility plots:\n",
      "<center><table><tr>\n",
      "<td>DBSA  <a href=\"files/experiment/Susceptibility_Dsilica_310K.JPG\"><img src=\"files/experiment/Susceptibility_Dsilica_310K.JPG\" width=\"250\"></a></td>  \n",
      "<td>HBSA  <a href=\"files/experiment/Susceptibility_HBSA_310K.JPG\"><img src=\"files/experiment/Susceptibility_HBSA_310K.JPG\" width=\"250\"></a></td>\n",
      "</tr></table></center>\n",
      "\n",
      "Comparison of MSD (label says 'Elastic Intensity' but it's wrong):  \n",
      "<center>SiO2_21A_3d_NoConnect.pdb  <a href=\"files/experiment/MSD_Dsilica_vs_HBSA_rebinned.jpg\"><img src=\"files/experiment/MSD_Dsilica_vs_HBSA_rebinned.jpg\" width=\"250\"></a></center>\n",
      "\n",
      "Maximum MSD corresponds to ~0.4A, very minimal flucuation. Thus, there is very little deuterated water surrounding the protein in the silica-encapsulated DBSA sample.  \n",
      "\n",
      "Slides that Panchao sent me providing more info on the samples:  \n",
      "\n",
      "<center><table><tr>\n",
      "<td><a href=\"files/experiment/sample_manifest.png\"><img src=\"files/experiment/sample_manifest.png\" width=\"300\"></a></td>  \n",
      "<td><a href=\"files/experiment/monomer_or_dimer.png\"><img src=\"files/experiment/monomer_or_dimer.png\" width=\"300\"></a></td>\n",
      "<td><a href=\"files/experiment/hydration.png\"><img src=\"files/experiment/hydration.png\" width=\"300\"></a></td>\n",
      "</tr></table></center>\n",
      "\n",
      "The last slide shows the relative weights of BSA (66054.4 amu per monomer), silica (60.08 amu per SiO2 molecule), and water (18.015 amu per molecule), corresponding to 1545 SiO2 and 622 water molecules for each BSA monomer."
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "<h2>Simulations</h2>"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "(<a href='#Simulations'>Top</a>)<a id='Simulations.references'></a><h3>References</h3>\n",
      "Some references regarding parameterization of force-fields describing amorphous silica\n",
      "- Review paper: [Ozboyaci16](files/simulation/references/Ozboyaci16.pdf) \n",
      "- [Cruz-Chu06](files/simulation/references/Cruz-Chu06.pdf)  \n",
      "- [Cruz-Chu09](files/simulation/references/Cruz-Chu09.pdf) \n",
      "- [Heinz13](files/simulation/references/Heinz13.pdf)  and [seminar presentation recording](files/simulation/references/Heinz13_presentation.wmv) and [slides](files/simulation/references/Heinz13_presentation_slides.pdf)\n",
      "- [Emami14](files/simulation/references/Emami14.pdf)  \n",
      "- [Emami14b](files/simulation/references/Emami14b.pdf)  \n",
      "- [Zheng16](files/simulation/references/Zheng16.pdf)\n",
      "- [Wensink00](files/simulation/references/Wensink00.pdf)\n",
      "\n",
      "References for mean square displacements in bulk materials containing silica  \n",
      "\n",
      "- [Fan17](files/simulation/references/Fan17.pdf)  \n",
      "- [Downs90](files/simulation/references/Downs90.pdf)  \n",
      "\n",
      "\n",
      "Simulations of proteins interacting with silica  \n",
      "- [Kubiak-Ossowska17](files/simulation/references/Kubiak-Ossowska17.pdf)  \n"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "(<a href='#Simulations'>Top</a>)<a id='Simulations.gromacs_tutorial'></a><h3>Gromacs tutorial for solvated Lyzosyme</h3>\n",
      "[Link](http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin/gmx-tutorials/lysozyme)  \n",
      "\n",
      ". <b>*.top</b>: (ASCII) The topology contains all the information necessary to define the molecule within a simulation. Thi- information includes nonbonded parameters (atom types and charges) as well as bonded parameters (bonds, angles, and dihedrals). Does not contain coordinates for the atoms.\n",
      "\n",
      "- <b>*.gro</b>: (ASCII) GROMACS-formatted structure file that contains residue and atom names of our system with the nomenclature of the specified force field. Also coordinates. \n",
      "\n",
      "- <b>*.itp</b>: (ASCII) Indicates a harmonic restraint of specified force constant to be applied for certain atoms of the  simulation.  \n",
      "\n",
      "- <b>*.mdp</b>: (ASCII) Run configuration parameters.\n",
      "\n",
      "- <b>*.tpr</b>: (binary) Run input. Integrates information from the *.top, *.gro, and *.mdp files.\n",
      "\n",
      "- <b>*.log</b>: (ASCII) standard output of the Gromacs run.\n",
      "\n",
      "- <b>*.edr</b>: (binary) values for various thermodynamic properties of the system versus time.\n",
      "\n",
      "- <b>*.trr</b>: (binary) trajectory (coordinates, and optionally velocities) versus time, in full-precision.\n",
      "\n",
      "- <b>*.cpt</b>: (binary) Checkpoint containing all the state variables, including coordinates and velocities, that allow one to continue a simulation.\n",
      "\n",
      "- <b>*.xvg</b>: (ASCII) Output of <code>gmx energy</code>, a file to be read by the plotting tool Xmgrace.\n",
      "\n",
      "- <b>*.ndx</b>: (ASCII) index containing user-defined sets of atoms.\n"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "(<a href='#Simulations'>Top</a>)<a id='Simulations.solvatedMBSA'></a><h3>Solvated Monomeric BSA box</h3>\n",
      "In subdirectory <i>simulations/monomer/solvatedBSA</i>.  \n",
      "\n",
      "<b>Obtaining a working initial PDB file</b>:  \n",
      "Structure [3v03.pdb](http://www.rcsb.org/pdb/explore.do?structureId=3V03) is a dimer. We select the second chain:\n",
      "<center><a href=\"files/simulation/monomer/solvatedBSA/3v03B_unclean.png\"><img src=\"files/simulation/monomer/solvatedBSA/3v03B_unclean.png\" width=\"250\"></a>  [3v03B_unclean.pdb](files/simulation/monomer/solvatedBSA/3v03B_unclean.pdb)</center>\n",
      "We clean up a bit the PDB file before obtaining the preprocessed Gromacs file.  \n",
      "\n",
      "- Removing crystallographic water Oxygen atoms.  \n",
      "- Removing residue \"ACT\".    \n",
      "- Submitting the PDB file to [pdb_hydro](http://lorentz.dynstr.pasteur.fr/pdb_hydro.php) server for generation of missing sidechains. We obtain structure [3v03B_noH.pdb](files/simulation/monomer/solvatedBSA/3v03B_noH.pdb) and log file [3v03B_PDB_Hydro.stdout](files/simulation/monomer/solvatedBSA/3v03B_PDB_Hydro.stdout). There are missing hydrogen atoms.\n",
      "- Removed residues \"CA\" with residue numbers 584 to 586. These are not part of the BSA protein.  \n",
      "- Added hydrogens with command:\n",
      "\n",
      "<code>gmx pdb2gmx -f 3v03B_noH.pdb -o 3v03B.pdb -p 3v03B.top -i 3v03B.itp -water spce -ff oplsaa &> 3v03B_pdb2gmx.log</code> \n",
      "\n",
      "to obtain [3v03B.pdb](files/simulation/monomer/solvatedBSA/3v03B.pdb), among other output files.  \n",
      "\n",
      "<b>Solvating the protein with water</b>:  \n",
      "Use OPLSAA force field and SPCE water model:  \n",
      "\n",
      "<code>gmx pdb2gmx -f 3v03B.pdb -o 3v03B.gro -p 3v03B.top -i 3v03B.itp -water spce -ff oplsaa &> 3v03B_pdb2gmx.log</code>\n",
      "\n",
      "outputs:  \n",
      "- 3v03B_pdb2gmx.log states 579 residues, 9173 atoms, and a total charge of -16e. We need to place 16 Na atoms in order to neutralize the system.\n",
      "- 3v03B.gro processed binary gromacs file.\n",
      "- 3v03B.itp positional restraints file.\n",
      "- 3v03B.top topology of the isolated molecule.\n",
      "\n",
      "The below command centers the protein in the box (-c), and places it at least 1.0 nm from the box edge (-d 1.0). The box type is defined as a cube (-bt cubic):  \n",
      "\n",
      "<code>gmx editconf -f 3v03B.gro -o 3v03B_newbox.gro -c -d 1.0 -bt cubic</code>\n",
      "\n",
      "Fill with solvent (water). Solvation is accomplished using solvate:  \n",
      "\n",
      "<code>cp 3v03B.top 3v03B_solv.top\n",
      "<code>gmx solvate -cp 3v03B_newbox.gro -cs spc216.gro -o 3v03B_solv.gro -p 3v03B_solv.top -scale 0.7</code>\n",
      "\n",
      "Add ions ([ions.mdp](files/simulation/monomer/solvatedBSA/ions.mdp)):  \n",
      "\n",
      "<code>cp 3v03B_solv.top 3v03B_solv_ions.top\n",
      "gmx grompp -f ions.mdp -c 3v03B_solv.gro -p 3v03B_solv_ions.top -o ions.tpr\n",
      "gmx genion -s ions.tpr -o 3v03B_solv_ions.gro -p 3v03B_solv_ions.top -neutral</code>  # select group number 13.  \n",
      "\n",
      "<b>Minimize the protein</b>:  \n",
      "(In subdirectory <i>simulation/monomer/solvatedBSA/minimize</i>)  \n",
      "Minimize options specified in file [minimize.mdp](files/simulation/monomer/solvatedBSA/minimize/minimize.mdp).\n",
      "<code>gmx grompp -f minimize.mdp -c 3v03B_solv_ions.gro -p 3v03B_solv_ions.top -o em.tpr</code> # creates binary input <code>em.tpr</code>.  \n",
      "<code>gmx mdrun -ntmpi 16 -ntomp 1 -pin on -deffnm em</code>  \n",
      "\n",
      "<b>Relax the water around the protein</b>:  \n",
      "(In subdirectory <i>simulation/monomer/solvatedBSA/minimize/relax</i>)  \n",
      "50ps NVT run using restraints on the protein atoms. Coordinates defining the equilibrium positions for the restraints are in file <code>3v03B.itp</code>. This file is encoded in the topology file <code>3v03B_solv_ions.top</code>. Options for the run specified in file [relax_1.mdp](files/simulation/monomer/solvatedBSA/minimize/relax/relax_1.mdp).  \n",
      "<code>gmx grompp -f relax_1.mdp -c em.gro -p 3v03B_solv_ions.top -o relax_1.tpr\n",
      "gmx mdrun -ntmpi 16 -ntomp 1 -pin on -deffnm relax_1</code>  # use 16 MPI threads, pin each thread to a diffeent core.  \n",
      "\n",
      "Continue with a 500ps NPT run using restraints on the protein atoms. The goal is to shrink the simulation box to tipycal densities. Input file is [relax_2.mdp](files/simulation/monomer/solvatedBSA/minimize/relax/relax_2.mdp).  \n",
      "<code>gmx grompp -f relax_2.mdp -c relax_1.gro -t relax_1.cpt -p 3v03B_solv_ions.top -o relax_2.tpr\n",
      "time gmx mdrun -ntmpi 16 -ntomp 1 -pin on -deffnm relax_2</code>  \n",
      "\n",
      "Recover some thermodynamics of the cell shrinkage:\n",
      "<code>gmx energy -f relax_2.edr -o relax_2.xvg</code>  \n",
      "\n",
      "<center><table><tr>\n",
      "<td><a href=\"files/simulation/monomer/solvatedBSA/minimize/relax/relax_2.png\"><img src=\"files/simulation/monomer/solvatedBSA/minimize/relax/relax_2.png\" width=\"300\"></a>  [relax_2.agr](files/simulation/monomer/solvatedBSA/minimize/relax/relax_2.agr)</td>  \n",
      "<td><a href=\"files/simulation/monomer/solvatedBSA/minimize/relax/relax_2.gif\"><img src=\"files/simulation/monomer/solvatedBSA/minimize/relax/relax_2.gif\" width=\"200\"></a>  [relax_2.gif](files/simulation/monomer/solvatedBSA/minimize/relax/relax_2.gif), [relax_2.vmd](files/simulation/monomer/solvatedBSA/minimize/relax/relax_2.vmd)</td>\n",
      "</tr></table></center>  \n",
      "\n",
      "<b>Center protein in the simulation box</b>:  \n",
      "<code>gmx trjconv -f relax_2.cpt -s relax_2.tpr -pbc mol -ur compact -center -o relax_2_centered.gro -boxcenter tric</code>  \n",
      "- <code>Select group for centering: Group     1 (        Protein) has  9173 elements</code>  \n",
      "- <code>Select group for output: Group     0 (         System) has 84738 elements</code>  \n",
      "Produces file <code>relax_2_centered.gro</code>.  \n",
      "<center><a href=\"files/simulation/monomer/solvatedBSA/minimize/relax/relax_2_centered.gif\"><img src=\"files/simulation/monomer/solvatedBSA/minimize/relax/relax_2_centered.gif\" width=\"200\"></a>  [relax_2_centered.gif](files/simulation/monomer/solvatedBSA/minimize/relax/relax_2_centered.gif)</center>\n",
      "\n",
      "<b>Select closest N water molecules to protein</b>:  \n",
      "The mass of the protein is 66054.497 a.m.u according to [3v03B_pdb2gmx.log](files/simulation/monomer/solvatedBSA/3v03B_pdb2gmx.log). Thus a hydration level $h=0.1$ corresponds to 6605.4497 a.m.u of water, corresponding to 6605.4497/18.01528 = 366 water molecules.  \n",
      "\n",
      "In the experiment, the estimated hydration level is 0.17. Thus, we have to include 1.7*366 = <b>622</b> molecules.  \n",
      "\n",
      "Save the following selection to file <i>selection_h0.17_waters.txt</i>:  \n",
      "\n",
      "<code>close_solvent_atoms = group \"SOL\" and within 0.1874 of group \"Protein\";\n",
      "close_solvent_molecules = same resindex as close_solvent_atoms;\n",
      "group \"Protein\" or close_solvent_molecules;</code>  \n",
      "\n",
      "We found a cutoff = 0.1874 to retain the 622 closest water molecules, simply by counting the number of lines containing \"SOL\" in file [stripped_system.pdb](files/simulation/monomer/solvatedBSA/minimize/relax/stripped_system.pdb) (see below).  \n",
      "\n",
      "Create index file <i>stripped_system_h0.17.ndx</i> with protein plus close waters, and also PDB file [stripped_system_h0.17.pdb](files/simulation/monomer/solvatedBSA/minimize/relax/stripped_system_h0.17.pdb)\n",
      "gmx select -f relax_2_centered.gro -s relax_2_centered.gro -sf selection_h0.17_waters.txt -on stripped_system_h0.17.ndx -ofpdb stripped_system_h0.17.pdb -pdbatoms selected  \n",
      "\n",
      "<center><a href=\"files/simulation/monomer/solvatedBSA/minimize/relax/stripped_system_h0.17.gif\"><img src=\"files/simulation/monomer/solvatedBSA/minimize/relax/stripped_system_h0.17.gif\" width=\"200\"></a></center>"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "(<a href='#Simulations'>Top</a>)<a id='Simulations.amorphous_from_MS'></a><h3>Amorphous silica box from Materials Studio post</h3>\n",
      "Issue [#4](https://github.com/jmborr/confinedBSA/issues/4)\n",
      "\n",
      "Silica $(SiO_2)_N$ has a sort-range order, whereby a Si atom is located at the center of a tetrahedron, whose vertices are defined by four O atoms.  \n",
      "\n",
      "In subdirectory <i>silica/amorphous_from_md/</i> and also Materials Studio project <i>silica</i> in virtual machine LEIOAVIRTUAL.  \n",
      "I downloaded 21A box from a [post](https://community.3dsbiovia.com/Communities_Topics?id=90650000000L9G5AAK&name=Materials+Studio&tid=09a500000004QfAAAU#/feedtype=SINGLE_QUESTION_DETAIL) in the Materials Studio community].  \n",
      "<center>SiO2_21A_3d_NoConnect.pdb  <a href=\"files/simulation/silica/amorphous_from_md/SiO2_21A_3d_NoConnect.png\"><img src=\"files/simulation/silica/amorphous_from_md/SiO2_21A_3d_NoConnect.png\" width=\"200\"></a></center>  \n",
      "Since the maximum diameter of BSA is ~77A, we use <i>SiO2_21A_3d_NoConnect.pdb</i> to create a 5x5x5 supercell of size ~107A. This precludes the protein from \"feeling itself\" via the periodic boundary conditions.  \n",
      "<center>SiO2_107A_3d_NoConnect.pdb  <a href=\"files/simulation/silica/amorphous_from_md/SiO2_107A_3d_NoConnect.png\"><img src=\"files/simulation/silica/amorphous_from_md/SiO2_107A_3d_NoConnect.png\" width=\"200\"></a></center>  \n",
      "\n",
      "<b>Inserting the protein in the silica box</b>:  \n",
      "\n",
      "We start with file <i>relax_2_centered.gro</i> from directory <i>simulation/monomer/solvatedBSA/minimize/relax</i>, which is contained in a cubic box of 9.41445nm. We have to insert the protein in the silica box <i>SiO2_107A_3d_NoConnect.pdb</i>, which is a cubic box of 10.6974nm. Thus, we resize the box of the protein plus water:  \n",
      "\n",
      "<code>gmx editconf -f relax_2_centered.gro -o relax_2_centered_resized.gro -bt cubic -box 10.6974</code>  \n",
      "\n",
      "<center><a href=\"files/simulation/silica/amorphous_from_md/confineBSA/relax_2_centered_resized.gif\"><img src=\"files/simulation/silica/amorphous_from_md/confineBSA/relax_2_centered_resized.gif\" width=\"200\">  [relax_2_centered_resized.gro](files/simulation/silica/amorphous_from_md/confineBSA/relax_2_centered_resized.gro)</a></center>  \n",
      "\n",
      "Next we select the 622 water molecules closest to the protein as we did before:  \n",
      "\n",
      "<code>gmx select -f relax_2_centered_resized.gro -s relax_2_centered_resized.gro -sf selection_h0.17_waters.txt -on stripped_system_h0.17.ndx -ofpdb stripped_system_h0.17.pdb -pdbatoms selected</code>  \n",
      "\n",
      "<center><a href=\"files/simulation/silica/amorphous_from_md/confineBSA/stripped_system_h0.17.gif\"><img src=\"files/simulation/silica/amorphous_from_md/confineBSA/stripped_system_h0.17.gif\" width=\"200\">  [stripped_system_h0.17.pdb](files/simulation/silica/amorphous_from_md/confineBSA/stripped_system_h0.17.pdb)</a></center>  \n",
      "\n",
      "We load <i>stripped_system_h0.17.pdb</i> and <i>SiO2_107A_3d_NoConnect.pdb</i> in VMD to check the protein is in the middle of the silica box:  \n",
      "\n",
      "<center><a href=\"files/simulation/silica/amorphous_from_md/confineBSA/overlap_protein_h0.17_and_silica.gif\"><img src=\"files/simulation/silica/amorphous_from_md/confineBSA/overlap_protein_h0.17_and_silica.gif\" width=\"200\"></a></center>  \n",
      "\n",
      "We need to \"clean up\" files <i>stripped_system_h0.17.pdb</i> and <i>SiO2_107A_3d_NoConnect.pdb</i> before merging protein and silica into a single PDB. We use [merge_silica_and_protein.py](files/simulation/silica/amorphous_from_md/confineBSA/merge_silica_and_protein.py) for this purpose\n",
      "\n",
      "<code>python merge_silica_and_protein.py</code> creates the following files:  \n",
      "\n",
      "- [stripped_clean_system_h0.17.pdb](files/simulation/silica/amorphous_from_md/confineBSA/stripped_clean_system_h0.17.pdb)  # modified protein file\n",
      "- [SiO2_107A_3d_clean.pdb](files/simulation/silica/amorphous_from_md/confineBSA/SiO2_107A_3d_clean.pdb)  # modified Silica file\n",
      "- [SiO2_protein_0.17.pdb](files/simulation/silica/amorphous_from_md/confineBSA/SiO2_protein_0.17.pdb) # protein and silica merged into a single PDB\n",
      "\n",
      "<b>Removing $SiO_2$ molecules overlapping the protein and water</b>  \n",
      "(tbd)  "
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "(<a href='#Simulations'>Top</a>)<a id='Simulations.cristobalite_2_2_2'></a><h3>Simulations of small cristobalite crystal</h3>\n",
      "\n",
      "In subdirectory <i>simulation/silica/cristobalite/cristobalite_2_2_2</i>\n",
      "\n",
      "With Materials Studio and []()\n",
      "<center><a href=\"files/simulation/silica/cristobalite/cristobalite_2_2_2/cristobalite_2_2_2.gif\"><img src=\"files/simulation/silica/cristobalite/cristobalite_2_2_2/cristobalite_2_2_2.gif\" width=\"150\"></a>  Small cristobalite crystal containing 96 atoms</center>  \n",
      "\n",
      "<b>Create topology with x2top utitily</b>\n",
      "\n",
      "We have modified oplsa.ff/ force field in <i>simulation/silica/cristobalite/cristobalite_2_2_2/oplsa.ff/</i> accounting for silicon type <code>SIO</code> and oxygen type <code>OSI</code> present in the crystal. We modified files <i>ffnonbonded.itp</i>, <i>ffbonded.ipt</i>, and <i>atomtypes.atp</i>.\n",
      "\n",
      "We create topology file with <code>gmx x2top</code>. This utitility <u>requires</u> description of connectivity between atom types in file <i>oplsa.f/atomname2type.n2t</i>.\n",
      "\n",
      "<code>gmx x2top -v -f cristobalite_2_2_2.gro -o cristobalite_2_2_2.top -ff oplsaa -name SIL -pbc -noparam</code>  \n",
      "\n",
      "After this, a minimization (minin) followed by equilibration at either constant volume (relax_1) or constant pressure (relax_2) are carried out. Run relax_2 is continued with another NPT run (relax_3) which can be used to gather data. Script [gmx_commands.sh](files/simulation/silica/cristobalite/cristobalite_2_2_2/gmx_commands.sh) containst the relevant gmx submissions.\n",
      "\n",
      "Relaxation at constant pressure [relax_2.mdp](files/simulation/silica/cristobalite/cristobalite_2_2_2/relax_2.mdp): \n",
      "\n",
      "- 0.1fs time step.\n",
      "- <code>verlet-buffer-tolerance=-1</code> to use <code>nstlist=1</code> and <code>rlist</code>.  \n",
      "- <code>tau_p=100</code> for slow relaxation of the volume\n",
      "\n",
      "Continuation at constant pressure [relax_3.mdp](files/simulation/silica/cristobalite/cristobalite_2_2_2/relax_3.mdp):\n",
      "- <code>gmx grompp -f relax_3.mdp -t relax_2.cpt -c relax_2.tpr -o relax_3.tpr</code>  # checkpoint required\n",
      "- 1fs time step\n",
      "- <code>verlet-buffer-tolerance=-1</code> to use <code>nstlist=10</code>  \n",
      "\n",
      "<center><a href=\"files/simulation/silica/cristobalite/cristobalite_2_2_2/relax_3_msd.png\"><img src=\"files/simulation/silica/cristobalite/cristobalite_2_2_2/relax_3_msd.png\" width=\"250\"></a> [relax_3_msd.agr](files/simulation/silica/cristobalite/cristobalite_2_2_2/relax_3_msd.agr)</center>\n"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "(<a href='#Simulations'>Top</a>)<a id='Simulations.cristobalite'></a><h3>Cristobalite</h3>\n",
      "\n",
      "From the [Crystallography Open Database](http://www.crystallography.net/cod/search.html) we search for the crystal structure of alpha-cristobalite, one of the morphologies of $SiO_2$:  \n",
      "\n",
      "<center>SiO2_107A_3d_NoConnect.pdb  <a href=\"files/simulation/silica/cristobalite/cristobalite_search.png\"><img src=\"files/simulation/silica/cristobalite/cristobalite_search.png\" width=\"300\"></a></center>  \n",
      "\n",
      "In Materials studio, the supercell replicating the unit cell 21 times along X and Y and 15 times along Z is able to surround BSA.  \n",
      "\n",
      "<center><a href=\"files/simulation/silica/cristobalite/cristobalite_21_21_15.gif\"><img src=\"files/simulation/silica/cristobalite/cristobalite_21_21_15.gif\" width=\"200\"></a>  [cristobalite_21_21_15.pdb](files/simulation/silica/cristobalite/cristobalite_21_21_15.pdb)</center>  \n",
      "\n",
      "A post in materials studio prescribes a recipe to build amorphous silica starting with the <code>amorphous cell</code> module. We can take parts of this recipe to create amorphous silica from a cristobalite supercell.  \n",
      "\n",
      "<pre>From: Jason DeJoannis\n",
      "I think we built that structure in partnership with Corning. It is documented in this paper:\n",
      "Vyas, Dickinson and Armstrong-Poston, Mol Sim, 32, 2, 135 (2006).\n",
      "\n",
      "Build bulk glass with Amorphous Cell\n",
      "Pack cell with SiO2 and Si2O4 fragments at density of 2g/mL\n",
      "cvff_aug forcefield with Ewald\n",
      "Anneal the bulk\n",
      "NVT at 7000K for 20ps\n",
      "NPT for 20ps\n",
      "Reduce T in 1000K increments with 20ps at each temperature down to 300K\n",
      "Create the surface\n",
      "Cleave and add vacuum slab 50-60A\n",
      "Extend surface supercell as needed\n",
      "Anneal the surface\n",
      "Start at 600K and cool to 300K in steps of 100</pre>"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "(<a href='#Simulations'>Top</a>)<a id='Simulations.confining_cristobalite'></a><h3>Protein confined in cristobalite</h3>\n",
      "\n",
      "In Materials studio, the supercell replicating the unit cell 21 times along X and Y and 15 times along Z is able to surround BSA.  \n",
      "\n",
      "<center><a href=\"files/simulation/silica/cristobalite/cristobalite_21_21_15.gif\"><img src=\"files/simulation/silica/cristobalite/cristobalite_21_21_15.gif\" width=\"200\"></a>  [cristobalite_21_21_15.pdb](files/simulation/silica/cristobalite/cristobalite_21_21_15.pdb)</center>  \n",
      "\n",
      "<b>Inserting the protein in the silica box</b>:  \n",
      "\n",
      "We start with file <i>relax_2_centered.gro</i> from directory <i>simulation/monomer/solvatedBSA/minimize/relax</i>, which is contained in a cubic box of 9.41445nm. We have to insert the protein in the cristobalite box <i>cristobalite_21_21_15.pdb</i>, which has dimensions 10.4406x10.4406x10.3835 nm. Thus, we resize the box of the protein plus water:  \n",
      "\n",
      "<code>gmx editconf -f relax_2_centered.gro -o relax_2_centered_resized.gro -bt triclinic -box 10.4406 10.4406 10.3835</code>  \n",
      "\n",
      "<center><a href=\"files/simulation/silica/cristobalite/confineBSA/relax_2_centered_resized.gif\"><img src=\"files/simulation/silica/cristobalite/confineBSA/relax_2_centered_resized.gif\" width=\"200\">  [relax_2_centered_resized.gro](files/simulation/silica/cristobalite/confineBSA/relax_2_centered_resized.gro)</a></center>  \n",
      "\n",
      "\n",
      "Next we select the 622 water molecules closest to the protein as we did before:  \n",
      "\n",
      "<code>gmx select -f relax_2_centered_resized.gro -s relax_2_centered_resized.gro -sf selection_h0.17_waters.txt -on stripped_system_h0.17.ndx -ofpdb stripped_system_h0.17.pdb -pdbatoms selected</code>  \n",
      "\n",
      "\n",
      "<center><a href=\"files/simulation/silica/cristobalite/confineBSA/stripped_system_h0.17.gif\"><img src=\"files/simulation/silica/cristobalite/confineBSA/stripped_system_h0.17.gif\" width=\"200\">  [stripped_system_h0.17.pdb](files/simulation/silica/cristobalite/confineBSA/stripped_system_h0.17.pdb)</a></center>  \n",
      "\n",
      "We need to \"clean up\" files <i>stripped_system_h0.17.pdb</i> and <i>cristobalite_21_21_15.pdb</i> before merging protein and silica into a single PDB. We use [merge_silica_and_protein.py](files/simulation/silica/cristobalite/confineBSA/merge_silica_and_protein.py) for this purpose.\n",
      "<code>python merge_silica_and_protein.py</code> creates the following files:  \n",
      "\n",
      "- [stripped_clean_system_h0.17.pdb](files/simulation/silica/cristobalite/confineBSA/stripped_clean_system_h0.17.pdb)  # modified protein file\n",
      "- [cristobalite_21_21_15_clean.pdb](files/simulation/silica/cristobalite/confineBSA/cristobalite_21_21_15_clean.pdb)  # modified Silica file\n",
      "- [SiO2_protein_0.17.pdb](files/simulation/silica/cristobalite/confineBSA/SiO2_protein_0.17.pdb) # protein and silica merged into a single PDB  \n",
      "\n",
      "<center><a href=\"files/simulation/silica/cristobalite/confineBSA/SiO2_protein_0.17.gif\"><img src=\"files/simulation/silica/cristobalite/confineBSA/SiO2_protein_0.17.gif\" width=\"200\"></a>  SiO2_protein_0.17.pdb</center>  \n",
      "\n",
      "<b>Removing $SiO_2$ molecules overlapping the protein and water</b>  \n",
      "\n",
      "The cristobalite lattice projected onto the XY plane can be viewed as a distorted squared lattice with the Si atoms occupying the corners, and the oxygen atoms located in the middle of the square sides. See below:\n",
      "\n",
      "<center><a href=\"files/simulation/silica/cristobalite/confineBSA/poretop/lattice_projected_Z.gif\"><img src=\"files/simulation/silica/cristobalite/confineBSA/poretop/lattice_projected_Z.gif\" width=\"200\"></a>  lattice_projected_Z.gif</center> \n",
      "\n",
      "With this arrangement, we can univoquely associate two oxygens to each Si, out of the four first-neighbors. We chose the oxygen with the largest X-component of its position relative to the Si, and the oxygen with the largest Y-component. This way we can paritition the cristobalite crystal in a unique way.\n",
      "\n",
      "Script [carve_silica.py](files/simulation/silica/cristobalite/confineBSA/poretop/carve_silica.py) uses [silicanet.py](files/simulation/silica/cristobalite/confineBSA/poretop/silicanet.py) takes the previous partition to create a hole in the cristobalite lattice. It takes as input file [SiO2_protein_0.17.pdb](files/simulation/silica/cristobalite/confineBSA/poretop/SiO2_protein_0.17.pdb) and partitions the cristobalite lattice into $SiO_2$ molecules. Molecules having any of its atoms overlapping the protein+water are removed. The overlapp parameter is tried with different values, from 1.0 to 2.5 Angstroms, producing different carved files. Also the names of the silica atoms will match those of the force field in subdirectory <i>simulation/silica/cristobalite/confineBSA/poretop/confinedBSA.ff/</i>:\n",
      "\n",
      "- [SiO2carved_ovl1.0_protein_0.17.pdb](files/simulation/silica/cristobalite/confineBSA/poretop/SiO2carved_ovl1.0_protein_0.17.pdb)  \n",
      "- [SiO2carved_ovl1.5_protein_0.17.pdb](files/simulation/silica/cristobalite/confineBSA/poretop/SiO2carved_ovl1.5_protein_0.17.pdb)  \n",
      "- [SiO2carved_ovl2.0_protein_0.17.pdb](files/simulation/silica/cristobalite/confineBSA/poretop/SiO2carved_ovl2.0_protein_0.17.pdb)  \n",
      "- [SiO2carved_ovl2.5_protein_0.17.pdb](files/simulation/silica/cristobalite/confineBSA/poretop/SiO2carved_ovl2.5_protein_0.17.pdb)  \n",
      "\n",
      "<center><a href=\"files/simulation/silica/cristobalite/confineBSA/poretop/SiO2carved_ovl1.5_protein_0.17.gif\"><img src=\"files/simulation/silica/cristobalite/confineBSA/poretop/SiO2carved_ovl1.5_protein_0.17.gif\" width=\"200\"></a>  SiO2carved_ovl1.5_protein_0.17.gif</center>  \n",
      "\n",
      "From the PDB files, it seems that overlapp > 1.5 leaves too many $SiO_2$ molecules and steric clashes that will not be resolved by a minimization run.  \n",
      "\n",
      "\n",
      "\n",
      "<b>Simulating the confining cristobalite without bonds</b>  \n",
      "As a first approximation, we assume each cristobalite atom is independent of the others, and it's subjected to a harmonic restraint. For cristobalite atoms in the interface of the protein+water, we can use as spring constant the value extracted from the experimental MSD of the dry protein (~2E-06 nm^2/K). For the bulk cristobalite atoms located away from the protein+water, we can use the spring constant derived from the simulations of the small cristobalite (~7E-07 nm^2/K).  \n",
      "\n",
      "in directory <i>simulation/silica/cristobalite/confineBSA/poretop/nobonds</i> we have modified force field <i>oplsaa.ff/</i> containing nonbonded interactions for the cristobalite atom types SIO, OSI, and OA. We modified the following files:  \n",
      "\n",
      "- [aminoacids.rtp](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/oplsaa.ff/aminoacids.rtp)\n",
      "- [atomname2type.n2t](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/oplsaa.ff/atomname2type.n2t)\n",
      "- [atomtypes.atp](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/oplsaa.ff/atomtypes.atp)\n",
      "- [ffnonbonded.itp](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/oplsaa.ff/ffnonbonded.itp)\n",
      "- [ions.itp](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/oplsaa.ff/ions.itp)\n",
      "- [residuetypes.dat](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/oplsaa.ff/residuetypes.dat)\n",
      "\n",
      "We use [SiO2carved_ovl1.5_protein_0.17.pdb](files/simulation/silica/cristobalite/confineBSA/poretop/SiO2carved_ovl1.5_protein_0.17.pdb) and create a suitable <i>.gro</i> file with script [adaptPDB.py](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/adaptPDB.py).\n",
      "\n",
      "<code>python adaptPDB.py</code> # produces [confinedBSA_0.gro](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/confinedBSA.gro)\n",
      "\n",
      "Create topology and restraint files:  \n",
      "\n",
      "<code>gmx pdb2gmx -f confinedBSA_0.gro -o confinedBSA.gro -p confinedBSA.top -i confinedBSA.itp -water spce -ff oplsaa &> confinedBSA.log </code>  \n",
      "\n",
      "produces the following files:  \n",
      "\n",
      "- [confinedBSA.gro](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/confinedBSA.gro) gro of whole system\n",
      "- [confinedBSA.itp](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/confinedBSA.itp) restraints for protein heavy atoms.\n",
      "- [confinedBSA_Other2.itp](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/confinedBSA_Other2.itp) topology of silica atoms\n",
      "- [confinedBSA_Other2_pr.itp](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/confinedBSA_Other2_pr.itp) restraints for silica atoms\n",
      "- [confinedBSA.pdb](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/confinedBSA.pdb) PDB of whole system\n",
      "- [confinedBSA_Protein.itp](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/confinedBSA_Protein.itp) topology of protein\n",
      "- [confinedBSA_Protein_pr.itp](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/confinedBSA_Protein_pr.itp) same as [confinedBSA.itp](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/confinedBSA.itp)\n",
      "- [confinedBSA.top](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/confinedBSA.top) topology of the system\n",
      "\n",
      "The total charge of the system is reported to be -342.900. To make the system neutral, we increase the positive charge of the SIO atoms. There are 23694 SIO atoms in the silica topology file [confinedBSA_Other2.itp](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/confinedBSA_Other2.itp). Thus, the new charge will be 1.1 + 342.9/23694 = 1.1144720182324641. We edit file [confinedBSA_Other2.itp](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/confinedBSA_Other2.itp) and make this change.\n",
      "\n",
      "We also updated the SIO charge in the following files of the force field:  \n",
      "\n",
      "- ions.itp\n",
      "- residues.rtp  (this is the file from which the silica charges are read by pdb2gmx)\n",
      "- ffnonbonded.itp\n",
      "\n",
      "We renamed Other2 to Silica, modifying [confinedBSA.top](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/confinedBSA.top) and producing producing new files [confinedBSA_Silica.itp](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/confinedBSA_Silica.itp) and [confinedBSA_Silica_pr.itp](files/simulation/silica/cristobalite/confineBSA/poretop/nobonds/confinedBSA_Silica_pr.itp)\n",
      "\n",
      "\n",
      "\n",
      "<b>Topology for the carved cristobalite</b>  \n",
      "\n",
      "script [evaluate_bonding.py](files/simulation/silica/cristobalite/confineBSA/poretop/evaluate_bonding.py) uses [silicanet.py](files/simulation/silica/cristobalite/confineBSA/poretop/silicanet.py) to calculate average bond and angle distances in the cyrstobalite using the cristobalite in [SiO2_protein_0.17.pdb](files/simulation/silica/cristobalite/confineBSA/poretop/SiO2_protein_0.17.pdb). The results:\n",
      "<code>bond_Si_O =  1.57833041748\n",
      "angle_O_Si_O =  109.149737561\n",
      "angle_Si_O_Si =  146.498743845</code>  \n",
      "\n",
      "Working directory: <i>simulation/silica/cristobalite/confineBSA/poretop/yesbonds</i>\n",
      "\n",
      "We use Zhang's force field with the previously found bond and angle average quantities. Resulting force field in subdirectory\n",
      "<i>simulation/silica/cristobalite/confineBSA/poretop/yesbonds/oplsaa.ff/</i>.\n",
      "\n",
      "From Simulating the confining cristobalite without bonds, we now that the charge of SIO should be <code>1.1144720182324641</code> to render a neutral silica+protein+water system. The charges are thus:\n",
      "\n",
      "- SIO +1.1144720182324641\n",
      "- OSI -0.55\n",
      "- OA -0.90\n",
      "- HO +0.40\n",
      "\n",
      "If we only want to simulate the silica, then a charge of SIO equal to <code>1.113796741791171</code> will render a neutral silica sytem.\n",
      "\n",
      "To create a topology file for the carved cristobalite (without the protein) we edit [SiO2carved_ovl1.5_protein_0.17.pdb](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/SiO2carved_ovl1.5_protein_0.17.pdb), removing the protein and water atoms, and then saving as\n",
      "[SiO2carved_ovl1.5.pdb](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/SiO2carved_ovl1.5.pdb).\n",
      "Then, script [topology_silica.py](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/topology_silica.py) will input this file and use module [silicatop.py](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/silicatop.py) to generate an include topology file. we produce \n",
      "[SiO2carved_ovl1.5_spw0.itp](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/SiO2carved_ovl1.5_spw0.itp) for a neutral silica+protein+water system, and [SiO2carved_ovl1.5_s0.itp](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/SiO2carved_ovl1.5_s0.itp) for a neutral silica system.\n",
      "\n",
      "We simulate the silica alone in directory <i>simulation/silica/cristobalite/confineBSA/poretop/yesbonds/onlysilica/</i> by copying [SiO2carved_ovl1.5_s0.itp](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/SiO2carved_ovl1.5_s0.itp) into [SIL.itp](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/onlysilica/SIL.itp). We had to tweak the last SIO with a slightly different charge (1.11505874179 instead of 1.11379674179) so that <code>gmx grommp</code> would report a neutral system. This is a numerical inaccuracy of <code>gmx grommp</code>. Also we had to comment the info in the <code>[ defaults ]</code> section because this info is previously set up in <i>oplsaa.ff/forcefield.itp</i>.\n",
      "\n",
      "With script [shift_indexes.py](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/onlysilica/shift_indexes.py) we shift the residue and atom indexes of [SiO2carved_ovl1.5.pdb](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/onlysilica/SiO2carved_ovl1.5.pdb) to create [confiningSIL.pdb](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/onlysilica/confiningSIL.pdb). This file starts with atom number one and has residue number equal to one.\n",
      "\n",
      "We manually created [confiningSIL.top](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/onlysilica/confiningSIL.top) and [minimize.mdp](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/onlysilica/minimize.mdp)\n",
      "\n",
      "<code>gmx editconf -f confiningSIL.pdb -o confiningSIL.gro\n",
      "gmx grommp -f minimize.mdp -c confiningSIL.gro -p confiningSIL.top -o minimize.tpr\n",
      "gmx mdrun  -ntmpi 16 -ntomp 1 -pin on -deffnm minimize</code>\n",
      "\n",
      "<center><a href=\"files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/onlysilica/minimize.gif\"><img src=\"files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/onlysilica/minimize.gif\" width=\"250\"></a>  [minimize.vmd](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/onlysilica/minimize.vmd) </center>  \n",
      "\n",
      "The minimization movie shows an expanding hole, causing a deformation in the crystal structure close to the hole surface. The expansion is due to the net negative charge caused by the dangling oxygens. These oxygens repel each other, and they quickly shield from each other by \"turning\" towards the bulk crystal. There are some $SiO_2$ molecules in the hole flying around. The expansion due to the negative charge will be less pronounced once we include the protein because it will act as a shielding insulator, but the steric clashes between protein and surface with induce an expansion on the hole.\n",
      "\n",
      "<b>Inserting the protein in the carved crystobalite with bonds</b>  \n",
      "\n",
      "In directory <i>simulation/silica/cristobalite/confineBSA/poretop/yesbonds/wholesystem/</i> we copied [SiO2carved_ovl1.5_spw0.itp](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/SiO2carved_ovl1.5_spw0.itp) into \n",
      "[SIL.itp](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/wholesystem/SIL.itp) and modified this file as follows:\n",
      "\n",
      "- commented out the <code>[ defaults ]</code> section\n",
      "- changed slightly the charge of the last SIO atom (from 1.11447201823 to 1.11468801823). <code>gmx grompp</code> complained the system had no net zero charge.\n",
      "\n",
      "We copied file [SiO2carved_ovl1.5_protein_0.17.pdb](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/SiO2carved_ovl1.5_protein_0.17.pdb)\n",
      "into [confinedBSA.pdb](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/wholesystem/confinedBSA.pdb)\n",
      "\n",
      "We created topology file [confinedBSA.top](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/wholesystem/confinedBSA.top) by hand, retrieving all info for the protein and water from previous topology file of the solvated BSA, [3v03B_solv_ions.top](files/simulation/monomer/solvatedBSA/3v03B_solv_ions.top)\n",
      "\n",
      "Input run files [minimize.mdp](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/wholesystem/minimize.mdp) and [relax_1.mdp](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/wholesystem/relax_1.mdp)\n",
      "\n",
      "<code>gmx editconf -f confinedBSA.pdb -o confinedBSA.gro\n",
      "gmx grompp -f minimize.mdp -c confinedBSA.gro -p confinedBSA.top -o minimize.tpr\n",
      "gmx mdrun  -ntmpi 16 -ntomp 1 -pin on -deffnm minimize\n",
      "gmx grompp -f relax_1.mdp -c minimize.gro -p confinedBSA.top -o relax_1.tpr\n",
      "gmx mdrun  -ntmpi 16 -ntomp 1 -pin on -deffnm relax_1  # 4ns in the NVT\n",
      "gmx grompp -f relax_2.mdp -c relax_1.gro -p confinedBSA.top -o relax_2.tpr\n",
      "gmx mdrun  -ntmpi 16 -ntomp 1 -pin on -deffnm relax_2  # 2ns in the NPT to relax the box size\n",
      "gmx grompp -c relax_2.tpr -t relax_2.cpt -p confinedBSA.top -f equil_1.mdp -o equil_1.tpr  # continuation but in the NVT\n",
      "gmx mdrun -ntmpi 16 -ntomp 1 -pin on -deffnm equil_1\n",
      "</code>\n",
      "\n",
      "<center><a href=\"files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/wholesystem/minimize.gif\"><img src=\"files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/wholesystem/minimize.gif\" width=\"250\"></a>  [minimize.vmd](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/onlysilica/minimize.vmd) </center>  \n",
      "\n",
      "The 2ns relaxation at NPT (<code>relax_2</code>) proved insufficiently long since the pressure became negative in the <code>equil_1</code> run. Thus, we extend the simulation for 10ns in Titan by appending 1ns jobs. For this we need to transfer to Titan all files <code>relax_2.*</code> and the following  \n",
      "\n",
      "- create soft link <code>ln -s relax_2.tpr relax_2.0.tpr</code>\n",
      "- template submission script [relax_2.extend.tpl](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/wholesystem/relax_2.extend.tpl)\n",
      "- substitution script: [relax_2.submitnext.sh](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/wholesystem/relax_2.submitnext.sh)\n",
      "\n",
      "In Titan, we just run the substitution script by passing as argument the number of the firs run:  \n",
      "<code>./relax_2.submitnext.sh 0</code>  \n",
      "Will trigger submission of the first nanosecond, followed by 9 more runs up to a total of 10ns\n",
      "\n",
      "After relaxation, we run an equilibrium run for 1ns in the NVT ensemble, and then append subsequent runs for a total of 12ns:\n",
      "\n",
      "- [equil_1.mdp](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/wholesystem/equil_1.mdp) and [equil_1.pbs](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/wholesystem/equil_1.pbs) for the initial 1n run.\n",
      "- [equil_1.extend.tpl](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/wholesystem/equil_1.extend.tpl) and [equil_1.submitnext.sh](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/wholesystem/equil_1.submitnext.sh) for the subsequent 11ns runs. We also need to create the soft link <code>ln -s equil_1.tpr equil_1.0.tpr</code> and then start the chained submission scripts with <code>./equil_1.submitnext.sh 0</code> .\n",
      "\n",
      "\n"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "(<a href='#Simulations'>Top</a>)<a id='Simulations.comparing_OA_ovl'></a><h3>Comparing confinements and OA charges</h3>\n",
      "\n",
      "We simulate four systems in order to find out which system provides us with the best comparison to experimental MSD at 300K:\n",
      "\n",
      "- $q_{OA}=0.90$ and silica-system overlap 1.5A\n",
      "- $q_{OA}=0.55$ and silica-system overlap 1.5A\n",
      "- $q_{OA}=0.55$ and silica-system overlap 2.5A\n",
      "- $q_{OA}=0.90$ and silica-system overlap 2.5A\n",
      "\n",
      "We already simulated the first system. The other three systems we create and simulate in subdirectories <i>simulation/confinedBSA/h017_q0.55_ovl1.5</i>, <i>simulation/confinedBSA/h017_q0.55_ovl2.5</i>, and <i>simulation/confinedBSA/h017_q0.90_ovl2.5</i>, respectively.\n",
      "\n",
      "With <code>gmx msd</code> we calculate MSD on the <code>equil_1.xtc</code> trajectories\n",
      "\n",
      "<center><a href=\"files/simulation/confinedBSA/analysis/comparison_MSD.png\"><img src=\"files/simulation/confinedBSA/analysis/comparison_MSD.png\" width=\"350\"></a>  [comparison_MSD.agr](files/simulation/confinedBSA/analysis/comparison_MSD.agr) </center>  \n",
      "\n",
      "The system $q_{OA}=0.90$ and silica-system overlap 2.5A approaches the most the experimental point. We select this system for the temperature scan.\n"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "(<a href='#Simulations'>Top</a>)<a id='Simulations.confining_cristobalite_q0.90ovl2.5'></a><h3>Protein confined in cristobalite with $q_{OA}=0.90$ and overlap=2.5A</h3>\n",
      "\n",
      "Files <i>confinedBSA.gro</i> and <i>confinedBSA.pdb</i> are same as for system $q_{OA}=0.55$ and overlap=2.5A. We use [topology_silica.py](files/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/topology_silica.py) to create [SiO2carved_q0.90_ovl2.5.itp](files/confinedBSA/simulation/silica/cristobalite/confineBSA/poretop/yesbonds/SiO2carve    .d_q0.90_ovl2.5.itp) which we rename to [SIL.itp](files/simulation/confinedBSA/h017_q0.90_ovl2.5/SIL.itp).\n",
      "\n",
      "As usual, we ran <code>gmx grompp -f minimize.mdp -c confinedBSA.gro -p confinedBSA.top -o minimize.tpr</code> and use the reported excess charge to edit file SIL.itp and there tune the charge of the SIO atoms to render a neutral system.\n",
      "\n",
      "We run the following jobs at 300K in Titan:\n",
      "\n",
      "- relaxation in the NVT ([relax_1.mdp](files/simulation/confinedBSA/h017_q0.90_ovl2.5/relax_1.mdp))\n",
      "- Extension of the previous relaxation ([relax_2.mdp](files/simulation/confinedBSA/h017_q0.90_ovl2.5/relax_2.mdp), [relax_2.pbs](files/simulation/confinedBSA/h017_q0.90_ovl2.5/relax_2.pbs)) followed by extension jobs ([relax_2.submitnext.sh](files/simulation/confinedBSA/h017_q0.90_ovl2.5/relax_2.submitnext.sh), [relax_2.extend.tpl](files/simulation/confinedBSA/h017_q0.90_ovl2.5/relax_2.extend.tpl))\n",
      "- Equilibrium job in the NVT ([equil_1.mdp](files/simulation/confinedBSA/h017_q0.90_ovl2.5/equil_1.mdp), [equil_1.pbs](files/simulation/confinedBSA/h017_q0.90_ovl2.5/equil_1.pbs))\n",
      "- Extension of the previous equililibrum job ([equil_1.submitnext.sh](files/simulation/confinedBSA/h017_q0.90_ovl2.5/equil_1.submitnext.sh), [equil_1.extend.tpl](files/simulation/confinedBSA/h017_q0.90_ovl2.5/equil_1.extend.tpl))\n",
      "\n",
      "Temperatures scan within subdirectory <i>simulation/confinedBSA/h017_q0.90_ovl2.5/Tscan</i>  \n",
      "\n",
      "We run the temperatures scan in Titan from $T_i=40$ to $T_e=420$ in increments of $20K$. Preparation of input files prior to submitting jobs to the queue done by running [bash_commands.sh](files/simulation/confinedBSA/h017_q0.90_ovl2.5/Tscan/bash_commands.sh).\n",
      "\n",
      "The following jobs:\n",
      "\n",
      "- Initial relaxation in the NPT ([relax_2.mdp.tpl](files/simulation/confinedBSA/h017_q0.90_ovl2.5/Tscan/relax_2.mdp.tpl), [relax_2.pbs](files/simulation/confinedBSA/h017_q0.90_ovl2.5/Tscan/relax_2.pbs))\n",
      "- Extension of the previous relaxation ([relax_2.extend.pbs.tpl](files/simulation/confinedBSA/h017_q0.90_ovl2.5/Tscan/relax_2.extend.pbs.tpl), [relax_2.submitnext.sh](files/simulation/confinedBSA/h017_q0.90_ovl2.5/Tscan/relax_2.submitnext.sh))\n",
      "- Equilibrium job in the NVT ([equil_1.mdp.tpl](files/simulation/confinedBSA/h017_q0.90_ovl2.5/Tscan/equil_1.mdp.tpl), [equil_1.pbs](files/simulation/confinedBSA/h017_q0.90_ovl2.5/Tscan/equil_1.pbs))\n",
      "- Extension of the previous equililibrum job ([equil_1.extend.pbs.tpl](files/simulation/confinedBSA/h017_q0.90_ovl2.5/Tscan/equil_1.extend.pbs.tpl), [equil_1.submitnext.sh](files/simulation/confinedBSA/h017_q0.90_ovl2.5/Tscan/equil_1.submitnext.sh))\n"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "(<a href='#Simulations'>Top</a>)<a id='Simulations.interfaceFF'></a><h3>The INTERFACE force field</h3>\n",
      "\n",
      "[https://bionanostructures.com/interface-md/](https://bionanostructures.com/interface-md/)  \n",
      "\n",
      "Latest version 1.5 in subdirectory <i>simulation/forcefields/INTERFACE_FF_1_5/FORCE_FIELDS</i>. Contains the force-field in PCFF and CHARMM format, plus a series of bulk and surface silica systems.\n",
      "\n",
      "- [INTERFACE_documentation_v1_5.doc](files/simulation/forcefields/INTERFACE_FF_1_5/INTERFACE_documentation_v1_5.doc)\n",
      "- [Description_of_silica_models.docx](files/simulation/forcefields/INTERFACE_FF_1_5/MODEL_DATABASE/SILICA/Description_of_silica_models.docx)\n",
      "\n",
      "\n"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "(<a href='#Simulations'>Top</a>)<a id='Simulations.ZhangFF'></a><h3>Zhang's force field</h3>\n",
      "\n",
      "Prof. Yingchun Liu sent me a force-field for silica after a personal request. This after reading Zheng16. This paper referenced Wensink00 paper.\n",
      "\n",
      "Located in subdirectory <i>simulation/forcefields/Zheng16</i>. Contains bonded and nonbonded parameters."
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "(<a href='#Simulations'>Top</a>)<a id='Simulations.protocol'></a><h3>Protocol for initial configuration for confinedBSA</h3>\n",
      "Issue [#6](https://github.com/jmborr/confinedBSA/issues/6)  \n",
      "\n",
      "A possible series of steps:\n",
      "- Solvate protein with water, and equilibrate the system. Center protein in the simulation box.   \n",
      "- Strip all but the closest $N$ water molecules, thus selecting a particular hydration level. This is the <i>hydratedBSA</i> system.  \n",
      "- Separately, simulate a system of silica to create amorphous silica. Dimensions of this system should be the same as the solvated equilibrated sytem. This annealing simulation entails melting the silica network followed by slow cooling.  \n",
      "- Overlap the hydratedBSA system onto the silica system, then remove overlapping $SiO_2$ molecules. This is the initial confinedBSA system. This system has a broken silica network due to unpaired electrons of the $Si$ atoms close to the water+protein.  \n",
      "- Fix the silica network with an annealing process in which the water+protein atoms are held frozen.  \n",
      "- Freeze the silica network and equilibrate the water + protein. This is the initial and equilibrated confinedBSA system.\n",
      "- Run equilibrium simulations of the confinedBSA system. The silica network is always kept frozen so simulations are in the NVT ensemble.  \n"
     ]
    },
    {
     "cell_type": "markdown",
     "metadata": {},
     "source": [
      "(<a href='#Table of Contents'>Top</a>)<a id='Syntax'></a><h3>HTML and Markdown Syntax Examples</h3>\n",
      "local link: [link](files/link)</br>\n",
      "remote link: <a href=\"http://ambermd.org/\">http://ambermd.org</a>\n",
      "<font face=\"courier new\"> font face=\"courier new\" </font><br/>\n",
      "$$S_{model}(Q,E)=A(Q)\\cdot S_{elastic}(E) + B(Q)\\cdot S_{simulation}(Q,E)\\otimes S_{elastic}(E) + C(Q)+D(Q)\\cdot E$$\n",
      "<pre> Quoted text </pre>\n",
      "<center><table><tr>\n",
      "<td><a href=\"files/image.png\"><img src=\"files/image.png\" width=\"300\" height=\"250\" alt=\"image here\"></a> <br/>\n",
      "    <i>image caption</i></td>\n",
      "<td>some text</td>\n",
      "</tr></table></center>"
     ]
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}