GROMACS 5.0.x CUDA/GPU Detection Failure With Ubuntu 14.04 nVidia 331.113 Update – Fix With An apt-get

If not for the near-20x speedup I’ve achieved running GROMACS on an nVidia GTX 770 Classified over an Intel i7 Extreme 6-core, nVidia in Ubuntu would almost be more trouble than its worth. The initial installation of the nVidia drivers from the nVidia website works, then the first time Ubuntu auto-updates the drivers to the latest-and-greatest, you’re never entirely sure what the next boot is going to look like – usually a black screen at best. And, if you found this page while looking for a solution to the nVidia driver update black/blank screen, my solution (which has worked without issue to date) is to ditch lightdm and use the GNOME Display Manager (gdm) instead (this apparently appears to be a theme with Ubuntu 14.04 installs on SSD drives as well).

sudo apt-get install gdm

Now, with that settled, the latest update (331.113) broke my GROMACS GPU install (performed using the steps outlined at: GROMACS 5.0.1, nVidia CUDA Toolkit, And FFTW3 Under Ubuntu 14.04 LTS (64-bit); The Virtues Of VirtualBox). The error for my system looks as follows:

Continue reading “GROMACS 5.0.x CUDA/GPU Detection Failure With Ubuntu 14.04 nVidia 331.113 Update – Fix With An apt-get”

GROMACS 5.0.1, nVidia CUDA Toolkit, And FFTW3 Under Ubuntu 14.04 LTS (64-bit); The Virtues Of VirtualBox

Summarized below are the catches and fixes from a recent effort to build GROMACS 5.0.1 with FFTW3 (single- and double-precision) and GPU support (so, single-precision). Also, a trick I’ve been doing with great success lately, using a virtual machine to keep my real machine as clean as possible.

0. The Virtues Of VirtualBox

Open source means never having to say you’re sorry.

I’ve made the above proclamation to anyone who’d listen lately who has any interest in using Linux software (because, regardless of what anyone says on the matter, it ain’t there yet as an operating system for general scientific users with general computing know-how). You will very likely find yourself stuck at a configure or make step in one or more prerequisite codes to some final build you’re trying to do, leaving yourself to google error messages to try to come up with some kind of solution. Invariably, you’ll try something that seems to work, only to find it doesn’t, potentially leaving a trail of orphaned files, version-breaking changes, and random downgrading only to find something else stupid (or not) fixed your build problems.

Continue reading “GROMACS 5.0.1, nVidia CUDA Toolkit, And FFTW3 Under Ubuntu 14.04 LTS (64-bit); The Virtues Of VirtualBox”

GROMACS 4.5.5, OpenMPI 1.6, And FFTW 3.3.2 Compilation Under Mountain Lion (OSX 10.8) With XCode (And A Little Help From Homebrew)

Minus a few glitches easily fixed with the right software, this build wasn’t bad at all (and thanks to Adam Lindsay for the title catch).

Now sitting in front of a new Core i7 MacBook Pro, one of the first compilations I wanted to have finished for new projects was GROMACS 4.5.5. As my procedure for compiling GROMACS 3.3.3 had been a highly-traveled page, I wanted to provide a brief summary of my successful 4.5.5 compilation.

A Few Piece Of Info

1. XCode

This used to be disc-download and install, now it’s available as a free download from the App Store (1.57 GB download, so plan to do something else while you wait for the download).

Continue reading “GROMACS 4.5.5, OpenMPI 1.6, And FFTW 3.3.2 Compilation Under Mountain Lion (OSX 10.8) With XCode (And A Little Help From Homebrew)”

Examining The Effects Of Vitamin B12 Conjugation On The Biological Activity Of Insulin: A Molecular Dynamic And In Vivo Oral Uptake Investigation

Published in MedChemComm (direct link: And Happy Belated New Year. After the methodological work that went into the Molecular Biosystems paper, this was a remarkably simple molecular dynamics study of the changes to vitamin B12 binding in transcobalamin II (TCII) with the B12 conjugated to the first amino acid side chain in the B-Chain of insulin. The structure of the B12-insulin conjugate is shown below in a molecular dynamics snapshot, which reveals that the binding of B12 to its TCII transport protein is negligibly affected.

And apparently the experiments went well, too. Cover hopefully to follow.

Susan Clardy-James, Damian G. Allis, Timothy J. Fairchild and Robert P. Doyle

Abstract: The practical use of the vitamin B12 uptake pathway to orally deliver peptides and proteins is much debated. To understand the full potential of the pathway however, a deeper understanding of the impact B12 conjugation has on peptides and proteins is needed. We previously reported an orally active B12 based insulin conjugate attached at LysB29 with hypoglycaemic properties in STZ diabetic rats. We are exploring an alternative attachment for B12 on insulin in an attempt to determine the effect B12 has on the protein biological activity. We describe herein the synthesis, characterization, and purification of a new B12-insulin conjugate, which is attached between the B12 ribose hydroxyl group and insulin PheB1. The hypoglycemic properties resulting from oral administration (gavage) of such a conjugate in STZ diabetic rats was similar to that noted in a conjugate covalently linked at insulin LysB2911, demonstrating the availability of both position on insulin for B12 attachment. A possible rationale for this result is put forward from MD simulations. We also conclude that there is a dose dependent response that can be observed for B12-insulin conjugates, with doses of conjugate greater than 10-9 M necessary to observe even low levels of glucose drop.

Vitamin B12 In Drug Delivery: Breaking Through The Barriers To A B12 Bioconjugate Pharmaceutical

In press in Expert Opinion On Drug Delivery (DOI:10.1517/17425247.2011.539200). The theory section (the only part I can properly speak to) builds on the discussion section of the full theory paper in Molecular Biosystems from earlier this year, providing an outlet for some of the more speculative design possibilities for trinary B12 bioconjugate design. Given that (1) there are mechanisms for cleavage at both of the proposed positions and (2) the molecular dynamics work indicates that, at least, TCII (transcobalamin II) can easily accommodate a bi-functionalized cobalamin, the A-B12-C design possibility is probably the most interesting long-term idea to come out of the computational side of the B12-insulin bioconjugate study (or so I argue).

Having “B12” and “cobalamin” in a blog post guarantees a bunch of useless moderation-necessary comments from vita-spam sites.

Susan M. Clardy, Damian G. Allis, Timothy J. Fairchild & Robert P. Doyle

Syracuse University, Syracuse, Department of Chemistry, NY 13244-4100, USA

Importance of the field: Vitamin B12 (B12) is a rare and vital micronutrient for which mammals have developed a complex and highly efficient dietary uptake system. This uptake pathway consists of a series of proteins and receptors, and has been utilized to deliver several bioactive and/or imaging molecules from 99mTc to insulin.

Areas covered in this review: The current field of B12-based drug delivery is reviewed, including recent highlights surrounding the very pathway itself.

What the reader will gain: Despite over 30 years of work, no B12-based drug delivery conjugate has reached the market-place, hampered by issues such as limited uptake capacity, gastrointestinal degradation of the conjugate or high background uptake by healthy tissues. Variability in dose response among individuals, especially across ageing populations and slow oral uptake (several hours), has also slowed development and interest.

Take home message: This review is intended to stress again the great potential, as yet not fully realized, for B12-based therapeutics, tumor imaging and oral drug delivery. This review discusses recent reports that demonstrate that the issues noted above can be overcome and need not be seen as negating the great potential of B12 in the drug delivery field.

The Binding Of Vitamin B12 To Transcobalamin(II); Structural Considerations For Bioconjugate Design – A Molecular Dynamics Study

In press, in the journal Molecular Biosystems. A first official foray into molecular dynamics-only (MD-only) computational work and I am pleased to report that the computational results not only make sense with respect to the experimental results, they also indicate a possible new way to use vitamin B12 for the oral delivery of bio-active molecules more complicated than the binary bioconjugates considered to date.

The Interesting Result

The conclusion from the previous study was that the insulin B Chain (figure below) acts as a tether to separate the structured region of insulin (the region with the largest inflexible steric bulk, see below) from the region of the transcobalamin II (TCII) that bind vitamin B12. It was then determined that the approach employed for the B12-insulin bioconjugate, simply linking one biomolecule onto another with known binding and transport properties (this is a common theme in all bioconjugate design), worked because the last 10 residues in the insulin B Chain (B22 to B30) are flexible in solution (they, in fact, cover the insulin binding region in the crystal form, then uncover this region in the biologically active form).

Continue reading “The Binding Of Vitamin B12 To Transcobalamin(II); Structural Considerations For Bioconjugate Design – A Molecular Dynamics Study”

The Vibrational Spectrum Of Parabanic Acid By Inelastic Neutron Scattering Spectroscopy And Simulation By Solid-State DFT

Available as an ASAP in The Journal of Physical Chemistry A. As a general rule in computational chemistry, the smaller the molecule, the harder it is to get right. As a brief summary, parabanic acid has several interesting properties of significance to computational chemists as both a model for other systems containing similar sub-structures and as a complicated little molecule in its own right.

1. The solid-state spectrum requires solid-state modeling. This should be of no surprise (see the figure below for the difference in solid-state (top) and isolated-molecule (bottom)). This task was undertaken with both DMol3 and Crystal06, with DMol3 calculations responsible for the majority of the analysis of this system (as has always been the case in the neutron studies reported on this site).

2. The agreement in the hydrogen-bonded N-H…O vibrations is, starting from the crystal structure, in poor agreement with experiment. You’ll note the region between 750 and 900 cm-1 is a little too high (and for clarification, the simulated spectrum is in red below). According to the kitchen sink that Matt threw at the structure, the problem is not the same anharmonicity one would acknowledge by Dr. Walnut’s “catalytic handwaving” approach to spectrum assignment (Dr. Walnut does not engage in this behavior, rather endeavors to find it in others where it should not be).

3. The local geometry of the hydrogen-bonding network in this molecular solid leads to notable changes in parabanic acid structure that, in turn, leads to the different behavior of the N-H…O vibrational motions. There is one potentially inflammatory comment in the Conclusions section that results from this identification. The parabanic acid molecule is, at its sub-structure, a set of three constrained peptide linkages that under go subtle but vibrationally-observable changes to their geometry because of crystal packing and intermolecular hydrogen bond formation. This means that the isolated molecule and solid-state forms are different and that peptide groups are influenced by neighboring interactions.

So, why should one care? Suppose one is parameterizing a biomolecular force field (CHARMM, AMBER, GROMOS, etc.) using bond lengths, bond angles, etc., for the amino acid geometry and vibrational data for some aspect of the force constant analysis. The structural data for these force fields often originates with solid-state studies (diffraction results). This means, to those very concerned with structural accuracy, that a geometry we know to be influenced by solid-state interactions is being used as the basis for molecular dynamics calculations that will NOT be used in their solid-state forms. Coupled with the different spectral properties due to intermolecular interactions, the description being used as the basis for the biomolecular force field likely being used in solution (solvent box approaches) is based on data in a phase where the structure and dynamics are altered from their less conformationally-restricted counterpart (in this case, solid-state).

A subtle point, but that’s where applied theoreticians do some of their best work.

Matthew R. Hudson, Damian G. Allis, and Bruce S. Hudson

Department of Chemistry, 1-014 Center for Science and Technology, Syracuse University, Syracuse, New York 13244-4100

Abstract: The incoherent inelastic neutron scattering spectrum of parabanic acid was measured and simulated using solid-state density functional theory (DFT). This molecule was previously the subject of low-temperature X-ray and neutron diffraction studies. While the simulated spectra from several density functionals account for relative intensities and factor group splitting regardless of functional choice, the hydrogen-bending vibrational energies for the out-of-plane modes are poorly described by all methods. The disagreement between calculated and observed out-of-plane hydrogen bending mode energies is examined along with geometry optimization differences of bond lengths, bond angles, and hydrogen-bonding interactions for different functionals. Neutron diffraction suggests nearly symmetric hydrogen atom positions in the crystalline solid for both heavy-atom and N-H bond distances but different hydrogen-bonding angles. The spectroscopic results suggest a significant factor group splitting for the out-of-plane bending motions associated with the hydrogen atoms (N-H) for both the symmetric and asymmetric bending modes, as is also supported by DFT simulations. The differences between the quality of the crystallographic and spectroscopic simulations by isolated-molecule DFT, cluster-based DFT (that account for only the hydrogen-bonding interactions around a single molecule), and solid-state DFT are considered in detail, with parabanic acid serving as an excellent case study due to its small size and the availability of high-quality structure data. These calculations show that hydrogen bonding results in a change in the bond distances and bond angles of parabanic acid from the free molecule values.

Crystal06 (v.1.0.2) And MPICH-1.2.7p1 In Ubuntu Desktop 8.10 (and 9.04, 64- and 32-bit) Using The Intel Fortran Compiler, Version 1.0

Update 19 May 2009 – This tutorial (and all subsequent modifications) are now on a separate page on this website and will not be modified further in this post.  This page is available HERE.  The forever-name PDF version of the tutorial is available here: crystal06_mpich_ubuntu_cluster.pdf

Pre-19 May 2009 – This document, the end of a very long and involved process, is available as a PDF download (for reading and printing ease) here: crystal06_mpich_ubuntu_cluster_V1.pdf


According the Crystal06 manual:

The CRYSTAL package performs ab initio calculations of the ground state energy, energy gradient, electronic wave function and properties of periodic systems. Hartree-Fock or Kohn-Sham Hamiltonians (that adopt an Exchange- Correlation potential following the postulates of Density-Functional theory) can be used. Systems periodic in 0 (molecules, 0D), 1 (polymers, 1D), 2 (slabs, 2D), and 3 dimensions (crystals, 3D) are treated on an equal footing. In each case the fundamental approximation made is the expansion of the single particle wave functions (‘Crystalline Orbital’, CO) as a linear combination of Bloch functions (BF) defined in terms of local functions (hereafter indicated as ‘Atomic Orbitals’, AOs).

Continue reading “Crystal06 (v.1.0.2) And MPICH-1.2.7p1 In Ubuntu Desktop 8.10 (and 9.04, 64- and 32-bit) Using The Intel Fortran Compiler, Version 1.0”

Building Parallel Abinit 5.6.x With OpenMPI 1.2.x (And NOT OpenMPI 1.3.x) From Sources In Ubuntu 8.x – iofn1.F90 Problem Solved

This post is an update to my previous post on building Abinit with OpenMPI in Ubuntu, with this post providing a workaround (solution?) to a run-benign but ultimately thoroughly aggravating issue with starting calculations in the abinip parallel build.

The description of the procedure, and the problem in the OpenMPI 1.3.x build, is as taken from the previous page (repeated so that the error makes its way and embeds itself a little deeper into the search engines).

To run parallel Abinit on a multi-processor box (that is, SMP.  The actual multi-node cluster setup is in progress), the command is SUPPOSED to be follows:

mpirun -np N /opt/etsf/abinit/5.6/bin/abinip < input.file >& output

Where N is the number of processors.  For mpirun, you need to specify the full path to the executable (which, for the build above, is as Abinit installs abinip when the build occurs in the /opt directory).  The input.file specification is as per the Abinit users manual so I won’t go into it here. You will also be asked to supply your password because I’ve done nothing to the setup of ssh (you are, in effect, logging into your machine to run the MPI calculation).

Continue reading “Building Parallel Abinit 5.6.x With OpenMPI 1.2.x (And NOT OpenMPI 1.3.x) From Sources In Ubuntu 8.x – iofn1.F90 Problem Solved”

New B12-Insulin-TCII-Insulin Receptor Cover Image For This Month’s ChemMedChem (March 2009)

As was the case for the first ChemMedChem December, 2007 cover issue (posted previously), the cover story in this month’s issue is a communication by myself and members and collaborators of the Robert Doyle Group here at Syracuse University.  In this case, the work for the cover image actually went into computational research published in the associated article (instead of just a pretty cover image to complement the associated article, which was the intent of the previous cover).

The image below shows the Transcobalamin II (TCII) protein (in teal ribbons, with a bound cyanocobalamin (B12) shown in red.  The PDB code for this complex is 2BB5) sitting within the surface-accessible fragment of the gigantic insulin receptor (PDB code 2DTG.  The cell membrane would be at the bottom of this image, with the remainder of the complete protein sitting both within the cell membrane and then into the cytoplasm).  Saving the lead-up to this structure generation for the associated published article, this image was created to show one of the most important steps in the Oral Insulin project being worked on in the Doyle Group, with the fact that we know it works making the validity of the image content all the more relevant.  In brief, this figure shows that the TCII/B12-Insulin complex can fit within the insulin receptor such that the insulin molecule can bind to its receptor position on the appropriately described insulin receptor (IR), thereby instigating the cascade of events that leads to cellular glucose uptake.

For a larger view, click on the image.

Continue reading “New B12-Insulin-TCII-Insulin Receptor Cover Image For This Month’s ChemMedChem (March 2009)”