Sunday, July 24, 2011

Blurring the boundary between linear scaling QM, QM/MM and polarizable force fields Part 2

My talk at WATOC 2011.  Slides can be found here
Here's how I made the video: I recorded my talk using the Voice Memo app on my iPhone.  Then I replayed the talk on my Mac using iTunes as I went through my slides on Powerpoint.  I recorded the screencast + audio using Screenflow.

Thanks to Anders Christensen for recording the video.

Thursday, July 21, 2011

Summarizing a paper using Prezi and Screencast-o-matic

In a previous post I summarized a paper using two programs: Prezi and Screencast-o-matic.  Both are free and easy to use.  The screencast above shows how I did it. Anyone can do this.

I did this on a mac, so I used the earphones/microphone that came with my iPhone.  The sound is not the greatest, but good enough I think.

The screencast is 9 minutes long, which might be too long.  It's hard to strike a balance between detail and the big picture.  Hopefully, I will improve with time.  Feedback is very welcome.

Update: be sure to check out the comments below

New paper: Ring current effects in proteins

Definitive Benchmark Study of Ring Current Effects on Amide Proton Chemical Shifts

Anders S. Christensen*, Stephan P. A. Sauer, and Jan H. Jensen*
Department of Chemistry, University of Copenhagen, Universitetsparken 5, DK-2100 Copenhagen Ø, Denmark
J. Chem. Theory Comput., 2011, 7 (7), pp 2078–2084
Abstract (the paper is summarized in the video at the end of the post)
The ring current effect on chemical shifts of amide protons (ΔδRC) is computed at the B3LYP/6-311++G(d,p)//B3LYP/aug-cc-pVTZ level of theory for 932 geometries of dimers of N-methylacetamide and aromatic amino acid side chains extracted from 21 different proteins. These ΔδRC values are scaled by 1.074, based on MP2/cc-pVQZ//B3LYP/aug-cc-pVTZ chemical shift calculations on four representative formamide/benzene dimers, and are judged to be accurate to within 0.1 ppm based on CCSD(T)/CBS//B3LYP/aug-cc-pVTZ calculations on formamide. The 932 scaled ΔδRC values are used to benchmark three empirical ring current models, including the Haigh–Mallion model used in the SPARTA, SHIFTX, and SHIFTS chemical shift prediction codes. Though the RMSDs for these three models are below 0.1 ppm, deviations up to 0.29 ppm are found, but these can be decreased to below 0.1 ppm by changing a single parameter. The simple point-dipole model is found to perform just as well as the more complicated Haigh–Mallion and Johnson–Bovey models.

The paper Instructions on how I made the video will appear in a future post. In the mean time, enjoy!