Fall ACS Boston

This fall's ACS meeting has been quite a different experience compared to the March ACS. There have been far more sessions dealing with computational drug discovery. In fact, my advisor organized a session on her U01 center, Community Structure-Activity Resource (CSAR), and the presented scoring challenge. It's not one of my projects in the lab, but it was really interesting to listen to some famous scientists in the field discuss what is needed for improving our ability to effectively score protein-ligand complexes. There were several community-wide discussion sessions related to opinions on pKa, changing our use of RMSD, the pressing need for high-quality data from a single source, and the impact of van der Waals on scoring. The work that groups across the country have performed with the CSAR data set will be published in a forthcoming special journal issue.

There were also a series of talks discussing fragment-based drug design, which has sky-rocketed in popularity recently. Fragments as a tool for developing lead compounds have been implemented into NMR, X-ray crystallography, rigid-protein surface mapping, as well as molecular dynamics-based methods. It has been an amazing opportunity for me to observe what other computational chemists regard as the primary challenges in our field, the different viewpoints they hold, and how they intend to solve them. It's also a lot less stressful not to be on the receiving end of that question!

Guest post: From the perspective of a more traditional chemist

This year’s spring meeting of the American Chemical Society presented an incredible opportunity for organic chemists of every level to hear talks from some giants in the field. Larry Overman, Mohammed Movassaghi, and Phil Baran all won awards this year, and their symposia were packed with lectures from the brightest minds in organic chemistry. I personally enjoyed Phil Baran’s talk the most: his chemistry is not only incredible, but he places molecules, reactions, and methodology in a historical context, which is not often seen in chemistry seminars. The ability of the Baran lab to construct the complex architecture of densely functionalized natural products in such short order is awe-inspiring. Larry Overman presented an overview of some of the reactions developed in his lab since the 70’s, most notably the Overman rearrangement. Donna Blackmond, now at Scripps, also gave an excellent talk discussing the work her lab has been performing in the investigation of the proline-catalyzed aldol reaction. In addition to the many fantastic organic seminars, there were also ample medicinal chemistry talks to attend. My personal interests steered me towards an afternoon session on photoaffinity labeling methods as a way to identify drug targets and binding sites. There was also an excellent series of talks discussing new ideas on modifying nuclear hormone receptor activity with novel ligands.

--Caleb Bates

ACS National Meeting Spring 2010

It’s easy for me, and for most other people I think, to become wrapped up in my own niche of research. I start to focus too much on getting to that next step, and forget about everything else that’s going on outside my desk. The ACS National Meeting Spring 2010 was the perfect reminder of all the other exciting research groups are doing.

The topic of this conference was “Chemistry for a Sustainable World”. But there were also a huge number of talks focused on computation and drug discovery. A couple of patterns jumped out from the presentation titles: there seems to be a growing interest in targeting allostery, in improving conformational sampling methods, applying fragment-based drug design, and calculation of binding free energies. Since my dissertation is focused on protein flexibility in drug design, there were a lot of people I was looking forward to hearing from. The ACS Award for Computers in Chemical and Pharmaceutical Research: Symposium in Honor of Kenneth M. Merz, Jr. had a number of excellent talks as part of the line-up.

William Swope from IBM gave a nice introductory discussion entitled “Sampling challenges in free energy calculations: identifying the boundaries of what is possible”. He reported on the outcome of a 2-year collaboration between GlaxoSmithKline and IBM to assess force fields for hydration free energies. One of the more interesting points he made was that “it doesn’t necessarily make sense to run all your simulations to the same run time” – a rather alien concept to me. I’m used to hearing that you can really only compare simulation results within the same time frame. But for the analysis of the impact of degrees of freedom on conformation, dipole moment and charge symmetrization of different small molecules, it kind of makes sense.

Another sampling presentation on atomistic simulations came from Vijay Pande at Stanford, who said that it “no longer has to be the case that full computational simulations will be as complicated as experimental simulations.” Victory for all the computational chemists out there who have to defend why their research is important, and why their research can’t simply be replaced by many experimentalists. Pande’s group is responsible for Folding@home, which has achieved a computing power of 8 petaflops!! Their research into protein folding has shown that raising the system temperature can actually slow protein folding down. His group has been studying adaptive sampling – using error from their model to point to states that need to be sampled more, adding in simulations on the fly, and bootstrapping. They’ve found that using a few short simulations can be more efficient that a long simulation of the same total time. This is the same technique that we use in the Carlson lab, with the same motivation – several unique, independent, shorter simulations of a system should yield similar if not superior information about the system compared to one very long simulation.

Ruben Abagyan gave a great talk on the challenges facing structure-based drug design that covered much of the recent studies published by his group. I am paraphrasing, but he noted that the SBDD challenges or “situations” are: protonation and refinement, cross-docking ligands to a single/multiple pockets with small induced fit, cross-docking ligands to pockets with large induced-fit, building dockable pocket models, selectivity profiling or repurposing, and an evolvable tech pipeline. He also did a great job of pointing out some of the problems affecting researchers’ understanding of crystal structures, to include interpreting electron density, potential fantasy heavy atoms, and ligands assigned an incorrect identity or given ambiguous placement.

Other great talks came from Adrian Roitberg: simulations with replica exchange should exchange as often as possible (but properly); Andrew McCammon: accelerated sampling; Matt Jacobson: understanding allostery through thermodynamic ensembles; Dagmar Ringe: using neutron beams to assign hydrogens in structures; and Richard Friesner: WaterMap – using MD to map out the free energy of water clusters in the active site (which is actually related to some of my research).

I had the chance to listen to two professors I’m interesting in post-docing for, and feel even more excited about their research after their talks than I did before.

I also had the great fortune to give a poster as part of the awards session for computational chemistry. It was really wonderful to have the chance to talk to other people about my current research. I hate sales, and when I was young and naïve I never expected to have to do any of it in research, but as it turns out – you do. Good research has a lot to do with being able to sell your vision: to funding agencies, to prospective employers/students/post-docs, and to your colleagues. No matter how great your idea may be, if you are afraid of public speaking, if you don’t show up prepared, if you sound bored – no one else is going to pay attention. Fortunately, I have had some time to get over being shy, and I am really excited about my current research project, which helps. I got a lot of great suggestions and was asked some interesting questions that I hadn’t considered before. I’m hoping to have a paper out on the research soon, and the opportunity to discuss and receive feedback as part of a poster session at ACS has really helped me to solidify my own vision of the project.

Hello

Welcome to the University of Michigan Brainworks: Medicinal Chemistry blog.