Saturday, April 29, 2017

Cheap but accurate calculation of chemical reaction rate constants from ab initio data, via system-specific, black-box force fields

Julien Steffen and Bernd Hartke 2017
Highlighted by Jan Jensen

Figure 1 from the paper. A flowchart of the EVB-QMDFF program implemented in this work, for the case of a DG-EVB-QMDFF calculation.

A few years ago I highlighted Grimme's General Quantum Mechanically Derived Force Field (QMDFF) - a black box approach that gives you a system-specific force field from a single QM Hessian calculation.  I missed the fact that Hartke and Grimme extended this approach to TSs using EVB, a year later. This EVB-QMDFF approach constructs EVB potentials connecting each pair of minima described by QMDFF.  To get the EVB parameters you need to supply the TS and 10-100 energies (and possibly 5-10 Hessian calculations) along the reaction path, depending on how complex an EVB potential is needed to describe the reaction.

What's the point of a system-specific reactive force field when you already have the TS and reaction path? Well, Steffen and Hartke show is that EVB-QMDFF can be used to perform the additional calculations needed for, for example, variational TS theory or ring polymer MD calculations to get more accurate rate constants.

Furthermore, just like for QMDFF for minima you could do all this for one conformation of ligands and use EVB-QMDFF for a conformer search or use the gas phase parameterized model to study the effect of explicit solvation.  It might even be possible to parameterize EVB-QMDFF using small ligands and then model the effect of larger ligands using the QMDFF parameters obtained for the minima.  However, all these potential uses still need to be tested.

I thank Jean-Philip Piquemal for bringing this paper to my attention

This work is licensed under a Creative Commons Attribution 4.0 International License.

Friday, April 28, 2017

Local Fitting of the Kohn−Sham Density in a Gaussian and Plane Waves Scheme for Large-Scale Density Functional Theory Simulations

Dorothea Golze, Marcella Iannuzzi, and Jürg Hutter (2017)
Highlighted by Michael Banck

Reprinted (adapted) with permission from Dorothea Golze, Marcella Iannuzzi, and Jürg Hutter. Journal of Chemical Theory and Computation, 2017 ASAP,  Copyright 2017 American Chemical Society.

Hutter et al. have published their LRIGPW (local resolution-of-the-identity gaussian and plane waves method) paper in JCTC. The image above taken from that paper highlights that much of the total runtime for conventional GPW (the main method implemented in the CP2K package) is spent on the description of the total charge density on real-space grids ("GPW grid", dark blue). Can you spot the orange bars for the same work done in the LRIGPW approach? This takes CP2K another big step forward.

Wednesday, April 19, 2017

New CCH contributors wanted

CCH has been been around for about five years and it's time to shake things up a bit in terms of how to contribute.

Who can contribute?
Anyone at any level of their career.  If you write a good highlight (in my opinion) then I'll post it (see below).  You don't have to be an expert in the subject of the highlight.  For example, I frequently highlight papers that use machine learning because I find it fascinating and want to learn about it, but I have yet to publish in that area.

In general, I would like CCH to be a diverse as possible in terms of subjects, career-stage, gender, geography, etc.  For example, the current highlights are mostly on small molecule, electronic structure-related papers and I'd love to have some more highlights on solid state and dynamics.  But more small molecule, electronic structure-related highlights are also fine.

How can I contribute?
If you are interested in contributing highlights to CCH send a highlight in any format to  If I like it, I'll post it. If you continue to send me highlights on a fairly regular basis (e.g. one every 2-3 months), then I'll add you to the editorial board and give you access to the site so you can post yourself.  If you make it on to the editorial board but don't post highlight  a 12 month period, I'll remove you from the editorial board again. This also goes for current editors, from today on.

Note that you don't commit yourself to be a regular contributor by sending a highlight. Contributing once is just fine.

You can also post a highlight on your own blog and send me a link for cross-posting.

What are the requirements for a highlight?
You can't highlight a paper you have co-authored.  The paper should be published within the current year, or last two years.  So in 2017, the paper must be published in 2015-2017.  You should highlight papers you (mostly) like and agree with - not papers of which you are highly critical.  You need at least a sentence or two on why you think the paper is interesting, and it should be of general interest to some fairly large subgroup of computational chemists. One thing I am not interested in is a steady stream of papers related to a very specific subject. It is fine to highlight a preprint that has been deposited but not accepted or published yet.

A bit about CCH
As mentioned above, CCH has been around for about five years.  CCH receives about 2200 page views per month, has about 450 Twitter followers, 1000 Facebook likes, and 330 followers on LinkedIn.

This work is licensed under a Creative Commons Attribution 4.0

Sunday, April 16, 2017

Quantifying Possible Routes for SpnF-Catalyzed Formal Diels–Alder Cycloaddition

Medvedev, M. G.; Zeifman, A. A.; Novikov, F. N.; Bushmarinov, I. S.; Stroganov, O. V.; Titov, I. Y.; Chilov, G. G.; Svitanko, I. V., J. Am. Chem. Soc. 2017, 139, 3942-3945
Contributed by Steven Bacharach
Reposted from Computational Organic Chemistry with permission

Medvedev, et al. have examined the cyclization step in the formation of Spinosyn A, which is catalyzed by the putative Diels-Alderase enzyme SpnF.1 This work follows on the computational study done by Houk, Singleton and co-workers,2 which I have discussed in this post (Dynamics in a reaction where a [6+4] and [4+2] cycloadditons compete). In fact, I recommend that you read the previous post before continuing on with this one. In summary, Houk, et al. found that a single transition state connects reactant 1 to both 2 and 3. The experimental product with the enzyme SpnF is 3. In the absence of enzyme, Houk, et al. suggest that reactions will cross the bispericyclic transition state TS-BPC (TS1 in the previous post) leading primarily to 2, which then undergoes a Cope rearrangement to get to product 3. Some molecules will follow pathways that go directly to 3.
The PCM(water)/M06-2x/6-31+G(d) study by Medvedev, et al. first identifies 560 conformations of 3. Next, they identified 384 TSs lying within 30 kcal mol-1 from the lowest TS. These can be classified as either TS-DA (leading directly to 3) or TS-BPC (which may lead to 2 by steepest descent, but can bifurcate towards 3). They opt to utilize the Atoms-in-Molecules theory to identify bond critical points to categorize these TS, and find that 144 are TS-BPC and 240 are TS-DA. (The transition state found by Houk, et al. is the second lowest energy TS found in this study, 0.29 kcal mol-1 higher in energy that the lowest TS and also of TS-BPC type.)

They also examined two alternative routes. First, they propose a path that first takes 1 to 4 via an alternative Diels-Alder reaction, and a second Cope rearrangement (TS-Cope2) takes this to 2, which can then convert to 3 via TS-Cope1. The other route involves a biradical pathway to either A or B. These alternatives prove to be non-competitive, with transition state energies significantly higher than either TS-DA or TS-BPC.

Returning to the set of TS-DA and TS-BPC transition states, while the former are more numerous, the latter are lower in energy. In summary, this study further complicates the complex situation presented by Houk, et. al. In the absence of catalyst, 1 can undergo either a Diels-Alder reaction to 3, or pass through a bispericyclic transition state that can lead to 3, but principally to 2 and then undergo a Cope rearrangement to get to 3. The question that ends my previous post on this subject — “ just what role does the enzyme SpnF play?” — remains to be answered.


1) Medvedev, M. G.; Zeifman, A. A.; Novikov, F. N.; Bushmarinov, I. S.; Stroganov, O. V.; Titov, I. Y.; Chilov, G. G.; Svitanko, I. V., "Quantifying Possible Routes for SpnF-Catalyzed Formal Diels–Alder Cycloaddition." J. Am. Chem. Soc. 2017, 139, 3942-3945, DOI: 10.1021/jacs.6b13243.
2) Patel, A.; Chen, Z.; Yang, Z.; Gutiérrez, O.; Liu, H.-w.; Houk, K. N.; Singleton, D. A., "Dynamically Complex [6+4] and [4+2] Cycloadditions in the Biosynthesis of Spinosyn A." J. Am. Chem. Soc. 2016, 138, 3631-3634, DOI: 10.1021/jacs.6b00017.


1: InChI=1S/C24H34O5/c1-3-21-15-12-17-23(27)19(2)22(26)16-10-7-9-14-20(25)13-8-5-4-6-11-18-24(28)29-21/h4-11,16,18-21,23,25,27H,3,12-15,17H2,1-2H3/b6-4+,8-5+,9-7+,16-10+,18-11+/t19-,20+,21-,23-/m0/s1
2: InChI=1S/C24H34O5/c1-3-19-8-6-10-22(26)15(2)23(27)20-12-11-17-14-18(25)13-16(17)7-4-5-9-21(20)24(28)29-19/h4-5,7,9,11-12,15-22,25-26H,3,6,8,10,13-14H2,1-2H3/b7-4-,9-5+,12-11+/t15-,16-,17-,18-,19+,20+,21-,22+/m1/s1
3: InChI=1S/C24H34O5/c1-3-19-5-4-6-22(26)15(2)23(27)11-10-20-16(9-12-24(28)29-19)7-8-17-13-18(25)14-21(17)20/h7-12,15-22,25-26H,3-6,13-14H2,1-2H3/b11-10+,12-9+/t15-,16+,17-,18-,19+,20-,21-,22+/m1/s1

This work is licensed under a Creative Commons Attribution-NoDerivs 3.0 Unported License.

Sunday, March 26, 2017

The Elephant in the Room of Density Functional Theory Calculations

Stig Rune Jensen, Santanu Saha, José A. Flores-Livas, William Huhn, Volker Blum, Stefan Goedecker, and Luca Frediani (2017) (Updated paywalled version)
Contributed by Jan Jensen

While basis set convergence sounds straightforward (though time-consuming) it is hard to rule out that underlying assumptions in  the design of the basis set influences the results.  However, converged basis set DFT results are needed to separate basis set errors from errors due to the functional. Multiwavelets, a systematic and adaptive multiresolution numerical solution of the one-electron problem, appear to be a way around this.

The paper presents PBE and PBE0 total energies, atomization energies, and dipoles moments for 211 molecules that are converged with respect to basis set to μHartree accuracy, and benchmarks Gaussian-type orbitals (GTOs), all-electron numeric atom-centered orbitals (NAOs) and full-potential augmented plane wave (APW) calculations. 

In the case of atomization energies, a quintuple GTO basis set (aug-cc-pV5Z) is needed to reach a 1 kcal/mol accuracy in both MAE and RMSE. For aug-cc-pVQZ the MAE is below 1 kcal/mol, but the RMSE is about 1.5 kcal/mol.  Perhaps more importantly, the maxAE goes from ca 10 to 2-5 kcal/mol on going from quadruple to pentuple basis set.  So even aug-cc-pV5Z cannot consistently reach the basis set limit for atomization energies!  It would have been very interesting to see whether extrapolated-CBS values are able to do this.

This dataset will be an important resource for developers of both DFT and basis sets.

This work is licensed under a Creative Commons Attribution 4.0

Simulation-Based Algorithm for Two-Dimensional Chemical Structure Diagram Generation of Complex Molecules and Ligand–Protein Interactions

Frączek, T. J. Chem. Inform. Model. 2016, 56, 2320-2335
Contributed by Steven Bacharach
Reposted from Computational Organic Chemistry with permission

Making a good drawing of a chemical structure can be a difficult task. One wants to prepare a drawing that provides a variety of different information in a clean and clear way. We tend to want equal bond lengths, angles that are representative of the atom’s hybridization, symmetrical rings, avoided bond crossings, and the absence of overlapping groups. These ideals may be difficult to manage. Sometimes we might also want to represent something about the actual 3-dimensional shape. So for example, the drawing on the left of Figure 1 properly represents the atom connectivity with no bond crossing, but the figure on the right is probably the image all organic chemists would want to see for cubans.

Figure 1. Two drawing of cubane

For another example, the drawing on the left of Figure 2 nicely captures the relative stereo relationships within D-glucose, but the drawing on the right adds in the fact that the cyclohexyl ring is in a chair conformation. Which drawing is better? Well, it likely is in the eye of the beholder, and the context of the chemistry at hand.
Figure 2. Two drawings of D-glucose.

Frączek has reported on an automated procedure for creating aesthetically pleasing 2-D drawings of chemical structures.1 The method involves optimizing distances between atoms projected onto a 2-D plane, along with rules to try to keep atom lengths and angles similar, and symmetrical rings, and minimize overlapping bonds. He shows a number of nice examples, especially of natural products, where his automated procedure PSM (physical simulation method) provides some very nice drawings, often noticeably superior to those generated by previously proposed schemes for preparing drawings.

Using the web site he has developed (, I recreated the structures of some of the molecules I have discussed in this blog. In Figure 3, these are shown side-by-side to my drawings. My drawings were generally done with MDL/Isis/Accelrys/Biovia Draw (available for free for academic users) with an eye towards representing what I think is a suitable view of the molecule based on what I am discussing in the blog post. For many molecules, PSM does a very nice job, sometimes better than what I have drawn, but in some cases PSM produces an inferior drawing. Nonetheless, creating nice chemical drawings can be tedious and PSM offers a rapid option, worthy of at least trying out. Ultimately, what we decide to draw and publish is often an aesthetic choice and each individual must decide on one’s own how best to present one’s work.

My Drawing
Figure 3. Comparison of my drawings vs. drawing made by PSM.


1) Frączek, T., "Simulation-Based Algorithm for Two-Dimensional Chemical Structure Diagram Generation of Complex Molecules and Ligand–Protein Interactions." J. Chem. Inform. Model. 2016, 56, 2320-2335, DOI: 10.1021/acs.jcim.6b00391.

This work is licensed under a Creative Commons Attribution-NoDerivs 3.0 Unported License.

Sunday, February 26, 2017

Towards full Quantum Mechanics based Protein-Ligand Binding Affinities

Stephan Ehrlich, Andreas H. Göller, and Stefan Grimme (2017)
Contributed by Jan Jensen

Erlich et al. presents absolute binding free energies for activated serine protease factor X (FXa) and tyrosine-protein kinase 2 predicted using DFT. Here I'll focus on FXa. The calculations are based on truncated model systems consisting of ca 1000 atoms. The geometries are optimised using HF-3c/C-PCM and select constraints, the RRHO free energy correction with DFTB3-D3, the electronic energy with PBE-3c, and the solvation free energy with COSMO-RS and PBE0/def-SVP. The energy terms are simply added together to give a total free energy and the binding free energy is simply the change in free energy upon binding without any additional corrections.

The MAD is similar to that found for host-guest complexes but there are clearly some outliers. The authors ascribe L19 and L27 to errors in the structures due to HF-3c artefacts, while L23 is ascribed to the movement of a crystal water molecule and L10 is the only charged ligand where the error in the solvation free energy is likely higher. The error is below 1.5 kcal/mol for 14 of the 25 ligands.

Clearly there is room for improvement but I do think the results are quite encouraging. A MM-PB(GB)SA study in which five different solvation models are tested for the same ligands found maximum $r$ values of 0.28 and 0.60 using ensemble averaged and energy minimised structures respectively. Furthermore, study determined the relative binding free energies using thermodynamic integration, which is generally considered the current gold standard in the drug design, for five ligand pairs (see table, energies in kcal/mol). Given that there is only five points any statistical analysis of the accuracy would be suspect, but I don't think TI can be said to outperform DFT.

DFT TI exp
5->18 4.0 -0.4 1.1
5->12 2.1 0.4 0.4
5->21 7.5 1.3 4.4
5->17 4.1 -0.2 -0.3
5->24 4.1 0.4 3.6

The real question is whether the DFT results can be systematically improved and the main sticking point here will ultimately be the solvation free energy, especially for charged ligands. The continuum model ultimately relies on a fit to experimental data so there is some degree of empiricism that is hard to remove. In principle it can be done by adding explicit water molecules but then the question is how to deal with the sampling in a cost effective way.

This work is licensed under a Creative Commons Attribution 4.0

Wednesday, February 22, 2017

Preparation of an ion with the highest calculated proton affinity: ortho-diethynylbenzene dianion

Poad, B. L. J.; Reed, N. D.; Hansen, C. S.; Trevitt, A. J.; Blanksby, S. J.; Mackay, E. G.; Sherburn, M. S.; Chan, B.; Radom, L., Chem. Sci. 2016, 7, 6245-6250
Contributed by Steven Bacharach
Reposted from Computational Organic Chemistry with permission

The new benchmark has been set for superbases. The previous record holder was LiO, with a computed proton affinity of 424.9 kcal mol-1. A new study by Poad, et al., examines the dianions of the three isomeric phenyldiacetylides: 1o1m, and 1p.1 Their computed proton affinities (G4(MP2)-6X) are 440.6, 427.0, and 425.6 kcal mol-1, respectively. The optimized geometries of these dianions are shown in Figure 1.



Figure 1. Optimized geometries of 1o1m, and 1p.

The authors also prepared these bases inside a mass spectrometer. All three deprotonate water, but do not deprotonate methane, though that might be a kinetic issue.
The authors speculate that 1o will be hard to beat as a base since loss of an electron is always a concern with small dianions.


1) Poad, B. L. J.; Reed, N. D.; Hansen, C. S.; Trevitt, A. J.; Blanksby, S. J.; Mackay, E. G.; Sherburn, M. S.; Chan, B.; Radom, L., "Preparation of an ion with the highest calculated proton affinity: ortho-diethynylbenzene dianion." Chem. Sci. 2016, 7, 6245-6250, DOI: 10.1039/C6SC01726F.


1o: InChI=1S/C10H4/c1-3-9-7-5-6-8-10(9)4-2/h5-8H/q-2
1m: InChI=1S/C10H4/c1-3-9-6-5-7-10(4-2)8-9/h5-8H/q-2
1p: InChI=1S/C10H4/c1-3-9-5-7-10(4-2)8-6-9/h5-8H/q-2

This work is licensed under a Creative Commons Attribution-NoDerivs 3.0 Unported License.

Sunday, February 12, 2017

Conformer-specific hydrogen atom tunnelling in trifluoromethylhydroxycarbene

Mardyukov, A.; Quanz, H.; Schreiner, P. R., Nat. Chem. 2017, 9, 71–76
Contributed by Steven Bacharach
Reposted from Computational Organic Chemistry with permission

The Schreiner group has again reported an amazing experimental and computational study demonstrating a fascinating quantum mechanical tunneling effect, this time for the trifluoromethylhydroxycarbene (CF3COH) 2.1 (I have made on a number of posts discussing a series of important studies in this field by Schreiner.) Carbene 2 is formed, in analogy to many other hydroxycarbenes, by flash vapor pyrolysis of the appropriate oxoacid 1 and capturing the products on a noble gas matrix.

Carbene 2t is observed by IR spectroscopy, and its structure is identified by comparison with the computed CCSD(T)/cc-pVTZ frequencies. When 2t is subjected to 465 nm light, the signals for 2t disappear within 30s, and two new species are observed. The first species is the cis conformer 2c, confirmed by comparison with its computed CCSD(T)/cc-pVTZ frequencies. This cis conformer remains even with continued photolysis. The other product is determined to be trifluoroacetaldehyde 3. Perhaps most interesting is that 2t will convert to 3 in the absence of light at temperatures between 3 and 30 K, with a half-life of about 144 h. There is little rate difference at these temperatures. These results are quite indicative of quantum mechanical tunneling.

To aid in confirming tunneling, they computed the potential energy surface at CCSD(T)/cc-pVTZ. The trans isomer is 0.8 kcal mol-1 lower in energy that the cis isomer, and this is much smaller than for other hydroxycarbenes they have examined. The rotational barrier TS1 between the two isomer is quite large, 26.4 kcal mol-1, precluding their interchange by classical means at matrix temperatures. The barrier for conversion of 2t to 3 (TS2) is also quite large, 30.7 kcal mol-1, and insurmountable at 10K by classical means. No transition state connecting 2c to 3 could be located. These geometries and energies are shown in Figure 1.





Figure 1. Optimized geometries at CCSD(T)/cc-pVTZ. Relative energies (kcal mol-1) of each species are listed as well.

WKB computations at M06-2X/6-311++G(d,p) predict a half-life of 172 h, in nice agreement with experiment. The computed half-life for deuterated 2t is 106 years, and the experiment on the deuterated analogue revealed no formation of deuterated 3.

The novel component of this study is that tunneling is conformationally selective. The CF3 group stabilizes the cis form probably through some weak HF interaction, so that the cis isomer can be observed, but no tunneling is observed from this isomer. Only the trans isomer has the migrating hydrogen atom properly arranged for a short hop over to the carbon, allowing the tunneling process to take place.


1) Mardyukov, A.; Quanz, H.; Schreiner, P. R., "Conformer-specific hydrogen atom tunnelling in trifluoromethylhydroxycarbene." Nat. Chem. 20179, 71–76, DOI: 10.1038/nchem.2609.


1: =1S/C3HF3O3/c4-3(5,6)1(7)2(8)9/h(H,8,9)
2: InChI=1S/C2HF3O/c3-2(4,5)1-6/h6H
3: InChI=1S/C2HF3O/c3-2(4,5)1-6/h1H

This work is licensed under a Creative Commons Attribution-NoDerivs 3.0 Unported License.