Norma research:howto

research:howto:amber99sb-ildn_tip4p-ew_and_namd

2011-07-14T10:57:06+02:00

AMBER, TIP4P-Ew and NAMD Consider this untested. Leap script: source leaprc.ff99SBildn loadamberparams frcmod.tip4pew loadamberparams frcmod.ionsjc_tip4pew loadoff tip4pewbox.off WAT = T4E set default FlexibleWater on lyta = loadPdb aligned.pdb check lyta solvateBox lyta TIP4PEWBOX 7.0 iso addions lyta Na+ 2 saveAmberParm lyta lyta.prmtop lyta.crd savepdb lyta lyta.pdb quit

research:howto:approaching_norma

2009-03-20T20:52:35+02:00

Approaching Norma The sooner you start viewing norma as a piece of laboratory equipment for which you will have to obtain proper training, the better. Previous experience with desktop computing machines does not carry-over to the cluster. You will have to shift mental gears in order to view the cluster for what really is: a multiuser, production-oriented, number-crunching computing machine, on which you have no physical access. Your aim should be to use the cluster, not to play with the clu…

research:howto:backup_backup_backup

2010-06-05T17:17:19+02:00

Method 1 : Optical media (DVD) If your files are less than 4.4 Gbytes (4,700,000 bytes) in total, doing a back-up using a DVD is a viable solution. To prepare a 'closed' DVD containing your data (and assuming that you've already loaded the DVD on norma), give

research:howto:calculate_and_plot_frequencies_of_transitions_between_conformers

2019-12-29T18:40:25+02:00

The idea is to use the information obtained from a cluster analysis performed with carma to calculate (a) the frequencies of transitions between distinct clusters, and, (b) to painlessly prepare a plot illustrating these transitions. The following is based on using the carma.clusters.dat file (or equivalent from grcarma) to (a) calculate the frequencies in the form of an adjacency matrix, and, (b) use the online server at to prepare a circos plot. The code (…

research:howto:calculation_of_a_principal_component_s_cosine_content

2009-03-09T12:57:15+02:00

Calculation of cosine content from principal components The equation to calculate cosine content from pricipal components as it appears in several papers (for example Proteins, 2007, 67, 569) contains a typographical error. The correct version looks like this:

research:howto:calculation_of_interaction_energies_using_namd

2009-02-24T19:41:01+02:00

Calculation of interaction energies You can do this nicely and easily using the NAMDenergy plugin of VMD. Or, you can do it directly with NAMD as follows: * Make a clean directory. * Copy to it the solute-only DCD & PSF files from your run (my.dcd and my.psf). * Copy to it the parameter file (par_all27_prot_na.inp). * Copy to it the solute-only PSF file from PSFGEN (my.pdb). * Edit the my.pdb file and make sure the occupancies the B-factor columns are both 1.0. Your file should lo…

research:howto:carma

2009-07-20T19:09:56+02:00

The examples that follow pertain to the latest version of carma (v.1.0b at the time of writing) and assume that you are using the GNU/Linux executable on a GNU/Linux machine. Furthermore, it will be assumed throughout that you are familiar with things like segment identifiers, atom identifiers, PSF file preparation, etc.

research:howto:carma_cookbook

2011-07-02T20:29:09+02:00

The carma cookbook For the examples that follow I will assume that you work on a GNU/linux box, that your favorite x-y graphing program is xmgr and that your postscript viewer is ghostscript. Getting rid of waters, ions, ..., and removing overall rotations-translations carma -v -w -fit -atmid ALLID -segid A my.dcd my.psf xmgr carma.fit-rms.dat mv carma.fitted.dcd protein.dcd mv carma.selected_atoms.psf protein.psf

research:howto:cluster_analysis_of_crossdcd_matrices_using_r

2013-11-28T16:43:54+02:00

Cluster analysis of molecular dynamics trajectories using RMSD matrices Calculation of the RMSD matrix using carma This is the preferred and possibly the easiest way. To calculate the RMSD matrix using only CA atoms, and every 500th structure give:

research:howto:comparing_two_secondary_structure_distributions

2017-09-29T21:31:55+02:00

Attempts to quantify how different (or similar) two trajectories are [1] eigenvector overlap This is an old and well-known method (there is a nefeli's wiki page discussing it here). This method won't do for highly disordered systems, for example, a folding simulation of a flexible peptide.

research:howto:comparison_of_carma-derived_clusters_and_rmsd_crossdcd_matrices

2011-05-01T16:40:10+02:00

Comparison between the carma-derived clusters and the RMSD (crossDCD) matrix Focus. The aim is to plot on the RMSD matrix the areas that correspond to a selected cluster obtained from carma (using either Cartesian or dihedral PCA). Proceed as follows:

research:howto:distance-matrix-based_clustering_of_pcs

2017-08-28T21:21:17+02:00

Distance-matrix-based clustering of principal components (dPCA or cPCA) The idea behind this method is to make an accurate dendrogram based on hierarchical clustering. For starters First of all, you need to have installed the R and some libraries. In the paradigm i will show you, i use the version 3.4.0 of R and the libraries (factoextra, FactoMineR, cluster and fpc). Second, depending on the physical memory that is available on your computer, the number of the frames that you will be able…

research:howto:estimating_minimal_pbc_box_for_a_folding_simulation

2017-02-28T12:19:53+02:00

Estimating the size of a safe minimal cubic PBC box for a peptide folding simulation Beware. What follows is primitive and possibly plain wrong. The idea is this : can we safely reduce the size of the cell for a folding simulation ? The word 'safely' in the previous sentence must be understood in a probabilistic sense : we want the smallest cubic cell for which the probability of neighboring peptide images interacting with each other (due to PBC) is less than an arbitrarily set threshol…

research:howto:guide_for_loading_data_to_dx

2012-01-27T19:06:54+02:00

Loading matrices to dx # cat > data.general file = /home/glykos/grid.matrix grid = 100 x 100 format = ascii interleaving = field majority = row field = field0 structure = scalar type = float dependency = positions positions = regular, regular, 0, 1, 0, 1 end

research:howto:hd_movies_using_pymol

2018-07-20T21:47:18+02:00

The following is not supported by the Educational version of PyMOL. If you installed PyMOL like that: sudo apt-get install pymol you are OK. The main steps are: 1. make a movie in PyMOL 2. type the commands that will make the final result high resolution 3. export frames as PNG files 4. stick the PNG files together in a movie

research:howto:high_dimensionality_cluster_analysis_based_on_pca

2012-01-25T18:52:08+02:00

High dimensionality PCA-based cluster analysis using k-means The idea is to automatically perform k-means-based cluster analysis using an arbitrary (large) number of principal components and a large (>50K) number of frames. See the Altis, Otten, Nguyen, Hegger & Stock (2008) paper for details. On the technical front, proceed as follows:

research:howto:high_dimensionality_cluster_analysis_with_snob

2012-01-31T17:26:08+02:00

High dimensionality PCA-based cluster analysis using SNOB Read first this page to get up to speed. Then come back. Welcome back. SNOB is a program written by Chris Wallace and encoding for the Minimum Message Length method for cluster analysis. Bibliography is available from this link. You can possibly start from this paper. On the technical front proceed as follows :

research:howto:histograms_of_principal_components

2011-06-19T18:49:03+02:00

One-dimensional histograms of selected principal components This is easy : awk '{print $2}' carma.dPCA.fluctuations.dat | plot -h awk '{print $3}' carma.dPCA.fluctuations.dat | plot -h awk '{print $4}' carma.dPCA.fluctuations.dat | plot -h awk '{print $5}' carma.dPCA.fluctuations.dat | plot -h awk '{print $6}' carma.dPCA.fluctuations.dat | plot -h ...

research:howto:md_and_nmr_calculation_of_chemical_shifts

2013-09-24T12:18:32+02:00

Calculation of chemical shifts from a molecular dynamics trajectory This will be based on the sparta+ program. The ad-hockery (calc_shifts) shown below is an attempt to automate the procedure, but does not cure the major problem which is this: too much I/O. For each frame, the dcd must be read, a pdb file must be written, the pdb file read by sparta+, and then everything should be processed and deleted. Which means that this is slow. This is very slow. This is so slow that it takes about 24 ho…

research:howto:md_and_nmr_calculation_of_j-couplings

2012-01-05T17:29:26+02:00

Calculation of J-couplings from molecular dynamics trajectories Coefficients for the various forms of Karplus equation come from Graf J, Nguyen PH, Stock G, Schwalbe H. (2007), with a correction for 3J(HN,C'). DFT1 and DFT2 variations come from Best RB, Buchete NV, Hummer G. (2008). These, in turn, come from Hu & Bax (1997), and Case, Scheurer, Brüschweiler (2000). See the source code at the end of this page for the actual equations used.

research:howto:md_and_nmr_calculation_of_noes

2010-04-19T17:10:27+02:00

MD and NMR, calculation of (r)^-3 and (r)^-6 averaged distances The assumption is that you have a set of experimental NOE-derived distances which you want to compare with the results from your simulation. For the discussion that follows you will need carma plus two extra programs: a perl script 'prep_proton.pl' and a C program 'noe_averaging'. Both of these are installed on norma, but for completeness, here is their source code.

research:howto:positional_autocorrelation_function

2009-04-03T20:39:48+02:00

Positional auto-correlation function The calculation can be performed with X-PLOR, but X-PLOR will not cooperate with DCD files containing unit cell information. You will have to use the old version of catdcd to prepare a new DCD file containing no unit cell information. Assuming that you did that, prepare an X-PLOR script similar to this

research:howto:preparation_of_diagrams_based_on_native_contacts_q_qs_s

2012-02-05T18:26:00+02:00

Calculation of native contacts and preparation of corresponding diagrams Recent versions of carma can calculate native contacts. The assumption is that the first frame of the trajectory is the 'native' structure (use catdcd to do this), and all calculations are performed with respect to this (first) frame. To perform the calculation using only Ca atoms try something like this :

research:howto:preparation_of_distance_vs._rg_plots

2009-03-09T13:02:56+02:00

Preparation of Rg vs N-C distance plots * Use paste to merge the two data files from carma (carma.Rgyration.dat AND carma.distances) so that each column is pasted next to the other * Use an editor to remove the columns that contain the frame number, or if you would rather use unix, tell it something like awk '{print $2 ” ” $4}' file_from_paste > new_file * Then, you need to make a grid out of the two columns in order to make a nice plot. The program grid will do the trick.

research:howto:preparation_of_energy_vs._surface_area_plots

2009-03-19T14:21:19+02:00

Energy vs. Surface area plots This is a variation of the distance vs. Rg plots. The difference is that the order parameters are surface and energy. To keep the calculation fast, the surface metric is the one calculated by carma (number of pixels making-up a fat molecular envelope). The energy comes from NAMD as described here.

research:howto:r_mds_and_friends

2017-12-02T21:07:08+02:00

R, MDS and friends

research:howto:rules_of_engagement

2009-03-20T19:16:27+02:00

Rules of engagement Thou shall not * Run jobs or do any serious calculation on the head node. That's why the compute-only nodes exist. * Never log-on directly to the compute nodes (i.e. you shant do anything like ssh n000?). Use the queuing system. If the queuing system doesn't work, it is a good opportunity to report it and get it fixed. * Do not assume that backing-up your data is someone else's responsibility. Because it isn't. * Do not leave your large files sitting on the clus…

research:howto:submitting_a_cns_job

2009-03-20T19:08:41+02:00

Submitting a CNS job CNS v.1.2 understands multithreading, and will automatically spawn four threads on a four-core machine. This means that you are better-off asking directly for -n4 cores (ie. a node) and avoid complications. To submit a job * Collect all your files in a directory. For example, a typical list for an annealing job would be :

research:howto:submitting_a_generic_serial_job_eg._ccp4

2009-03-26T15:04:00+02:00

Submitting a generic serial job If your job is neither parallel, nor multithreaded (as is, for example, all of CCP4 programs including refmac), then your program will just 'stick' to one and only one core and nothing much is to be gained from the point of view of performance. In reality it is worse than that: your single-threaded job (if submitted to the fast queue), may be resposible for practically stopping the execution of a full scale parallel job. So, be a good citizen and do not submit a…

research:howto:submitting_a_mpi_parallel_job

2009-11-16T19:08:38+02:00

Submitting an MPI (parallel) job If you already have a properly built and tested executable, submitting the job is a piece of cake. For example, to submit a parallel Qs job, proceed as follows: * Prepare a submission script with a suitable name, eg. Qs.csh, containing the following (the -np 4 flag says that we want to use four cores for this job) :

research:howto:submitting_a_namd_job

2011-09-28T16:14:28+02:00

Submitting an NAMD job to Norma * Transfer all files needed for your job to Norma using ftp or sftp. * Organise your files in a sensible way (use mkdir). * If you working with a completely new system (see the benchmarks page), consider the possibility of tuning the simulation parameters for Norma. If you do not know how, ask for advice. * cd to the directory containing the NAMD files and make sure that everything you need for the run is there. * Submitting the job to the queue using t…

research:howto:submitting_a_vina_autodock_job

2014-03-20T16:43:36+02:00

Submitting a vina (autodock) job Vina understands multithreading, and will automatically spawn four threads on a four-core machine. This means that you are better-off asking directly for -N1 (ie. one node) and avoid complications. To submit a job * Collect all your files in a directory. For example :

research:howto:torsion-rmsd_calculation_and_applications

2017-10-02T12:27:32+02:00

Torsion-RMSD, calculation and applications Following the idea of dihedral PCA, instead of calculating the Cartesian-based RMSD between two structures, you can calculate the RMSD between their torsion angles (possibly including the χ1 angles as well). As with dPCA, the nice thing about this t-RMSD is that you do not have to least-squares superimpose the structures. This in turn makes the method suitable even for highly flexible systems.

research:howto:using_a_4fs_timestep_through_hydrogen_mass_repartitioning

2017-10-12T12:51:35+02:00

Using a 4fs timestep through Hydrogen Mass Repartitioning The latest (?) paper on the subject is this, but see also some well-thought criticism here. Part I : Technical stuff Assuming that you use the AMBER force field, preparing a simulation using HMR is a piece of cake : the implementation has been completely automated through the program 'parmed'. It goes like this :

research:howto:using_methanol_or_tfe_as_solvent

2018-11-22T10:57:44+02:00

Using methanol, DMSO, TFE or water/TFE mixtures as solvents in folding simulations This page discusses the technical procedures required to run a folding molecular dynamics simulation (using the AMBER family of force fields) in methanol or TFE. In both cases, we examine the case of homogeneous solvent as well as a mixture with water). If you wanted to do a proper job (and not this hack) you can take ideas from here.

research:howto:using_the_amber12sb_and_amber14sb_forcefield

2017-09-23T11:03:08+02:00

AMBER ff12SB Make sure that AMBERHOME points to /usr/local/amber12. # # cat leap.script source leaprc.ff12SB loadamberparams frcmod.ionsjc_tip3p test = loadPdb aligned.pdb check test solvateBox test TIP3PBOX 7.0 iso addions test Na+ 2 saveAmberParm test test.prmtop test.crd savepdb test test.pdb # # # tleap -s -f leap.script # # vmd -dispdev text test.prmtop test.crd set all [atomselect top all] $all writepsf ionized.psf quit # #

research:howto:using_the_amber12sb_forcefield

2012-08-08T18:06:04+02:00

research:howto:using_the_amber15sb_tip3p-fb_forcefield

2018-12-16T11:31:19+02:00

AMBER ff15SB plus TIP3P-FB water model Make sure that AMBERHOME points to /usr/local/amber16 : # # cat leap.script source leaprc.protein.fb15 source leaprc.water.fb3 fsd1 = loadPdb aligned.pdb check fsd1 solvateBox fsd1 TIP3PFBOX 9.5 iso addions fsd1 Cl- 5 saveAmberParm fsd1 fsd1.prmtop fsd1.crd savepdb fsd1 fsd1.pdb # # tleap -s -f leap.script #

research:howto:using_the_amber_forcefield

2013-07-11T19:20:38+02:00

Using the AMBER forcefield with NAMD In our lab the standard protocol for running MD simulations entails the NAMD with the CHARMM forcefield. So, every time we wish to push our lack with some other widespread and well-known forcefield, is mainly for comparison reasons. Using the OPLS-AA is quite straightforward and described here Using the OPLS-AA forcefield Using the AMBER forcefield is more tricky especially if you want to use it with the NAMD engine. AMBER11 is not free, but the AMBERTOO…

research:howto:using_the_charmm_forcefield_with_the_cmap_correction

2009-06-19T18:28:00+02:00

Using the CMAP correction with the CHARMM forcefield The only difference with the standard procedure is the topology and parameter files needed. Nothing else in the various scripts needs changing. The files you need are /usr/local/toppar/par_all22_prot.inp /usr/local/toppar/top_all22_prot.inp

research:howto:using_the_opls-aa_forcefield

2009-06-19T18:41:57+02:00

Using the OPLS-AA forcefield with NAMD Assuming that your system comprises only protein, TIP3 waters, sodium and chloride ions, proceed as follows: Make /usr/local/toppar/non_charmm your working directory. Locate the files named top_opls_aa.inp and par_opls_aa_modified.inp. These are the topology and parameters files that you will have to use for your simulation (the parameter file has been modified to include parameters for the sodium and chloride ions, which explains the modified suffix).