Hex is an interactive protein docking
and molecular superposition program,
written by
Dave Ritchie.
Hex understands protein and DNA structures in
Hex will run on most Windows-XP, Linux and Mac OS X PCs. The recent versions now include CUDA support for Nvidia GPUs. On a modern workstation, docking times range from a few minutes or less when the search is constrained to known binding sites, to about half an hour for a blind global search (or just a few seconds with CUDA). On multi-processor Linux systems, docking calculation times can be reduced in almost direct proportion to the number of CPUs and GPUs used. The calculations can be accelerated by using an optional disc cache (strongly recommended) of about 1 GB of disc space.
For Academic and Governmental users, a runtime version of Hex may be downloaded free of charge. You are asked only to comply with a wordy, but mostly harmless, licence agreement. Corporate users may also download the software for evaluation, although prior agreement with the author is required if the program is to be used for profit.
The User Manual is included in the software installation, and is also available here:
The basic installation includes some examples. A more extensive set of docking test cases can be obtained from Zhiping Weng's Docking Benchmark.
If you wish to try Hex without downloading the actual program, please
use the link below to access the on-line docking server.
With our new GPU-based code,
a typical docking job takes about 5 minutes
15 seconds on our server.
The "Sam" (symmetry assembler) program uses the Hex polar Fourier correlation technique to build symmetrical protein complexes with any of the naturally occurring point group symmetries (Cn, Dn, T, O, and I). Check it out!
KBDOCK is a database of all known structural domain-domain interaction, built directly from the PDB and Pfam (2013 snapshots). If structural homologues exist for your docking target, KBDOCK might be able to help! If KBDOCK finds homology templates, it can also be used as a handy way to launch a docking job with HexServer.
The GPU accelerated version of Hex was developed using CUDA with the help of a kind donation of a GeForce GTX 9800 card from the Nvidia Professor Partnership Program. Here are some further links relating to CUDA.
So how good is Hex, or any other docking algorithm for that matter? The CAPRI (Critical Assessment of Prediction of Interactions) experiment aims to answer that question. Since the summer of 2001, Hex has taken part in all of the blind prediction rounds of CAPRI, in which the task was to predict the structures of protein complexes whose structures were in the process of being solved by X-ray crystallography. There are now two special editions of Proteins which describe the algorithms and results: Vol 52(1), July 2003, and Vol 60(2), Aug 2005. Overall, in Rounds 1-2 Hex did quite well, scoring 2 close hits for two of the seven targets. In Rounds 3-5, Hex scored a 1.8A RMS hit at rank 6 for target 12 (cohesin/dockerin) and got some further low to medium accuracy hits for targets 10, 11, 12 and 13, but missed the rest!
There is now an on-going Call For Targets for further prediction experiments. If you are a crystallographer in the process of solving the structure of a protein (or DNA) complex, please consider submitting it as a CAPRI target!
Planet Hex has been orbiting somewhere between the earth and the moon since 1999. Have you discovered planet Hex yet?
Hex moved to France in December 2008. The fact that French people often describe France as "Le Hexagon" is a complete coincidence (but a nice one).
The following publications describe the algorithms and results obtained with Hex:
For a review about the general field of protein docking, please see:
Much of the early development of Hex was done during projects funded by the BBSRC. Currently, the main sources of support are the ANR which funds some of my projects, and INRIA which pays my salary.