if someone wants to develop an application outside of WCG, then they have every opportunity to do so. But if that is the idea, it need not be promoted inside WCG. That would be (is) confusing that somehow this new project will benefit or be supported by WCG.

WCG is clearly not going to support an open source application development project, but that should not be a deterrent to anyone who wants to develop an open source DC project. You can even do so for Cancer or Aids, but it won't be hosted on the WCG servers.

SETI is one of many separate applications that have no relationship with WCG, for example. Similarly, if SETI ever supports GPUs is meaningless to the WCG applications. WCG is also not impacted if the SETI application is open source or proprietary.

I have pointed some old threads to this one, so that ppeople who actually want to contribute to native x64/GPU development will start doing something in this direction.

Lets leave WCG people doing their great job, and try ourselves to setup open project for the native x64/GPU science application.

Never saw this in the small print before but next AutoDock deployment at WCG will make use of the Vina version.

Hello Steffen, Sekrob and all other respected contributors,

Lets make sure our goal is clear. I understand the goal is a creation of open source, grid enabled molecular docking software that would be optimized with time to include new CPU technologies (64 bit and SSE) and GPU technologies.

Resulting software can be used directly by scientists or be a base for future WCG projects, possibly leading to closed source deployments.

It is great news that Vina is going to be used in new FAH iteration. It is faster and tecnically cleaner than autodock. Autodock is difficult to port out of the box due to some library dependencies.

The startup tasks are:
- create project environment such as source code repository, project web page, email list,
- get in touch with WCG/Scripps and ask if they actually want some improvements in Vina that would be of direct value to them that might be performed on our project,
- with time get in touch with some research institutes and see which of them might be interested in using it.

Do you think it is a good start?

Hello,



from what I see, at least the 64 bit (and sse with it, although not manually optimised for it) version of AutoDock is already there, but just like the translations it is just not redistributed. My aim is hence to have a platform that would allow to integrate such tools earlier - and invalidate all results of those platforms if those are worse than what versions yield or incorrect. So: platform first, technologies to plug in then are already available or come over time.



Yes, and if they make our open effort redundant, then this is an achievement. But I would particularly be happy to learn about other projects also outside the WCG reusing the Open infrastructure. But given that the WCG is a good partner for what they are doing, we should be very happy for them to exist and take extra efforts from the scientists away. It is just that we think that they could do more than they are doing and they say they don't have the resources to go for it. If we had an Open Source WCG infrastructure, then we would just extend that, of course, but we don't have that and as such the idea for an Open Source alternative arose.



Well, I don't know for Vina (which is not open source to my knowledge), but at least for Linux platforms, there is no difficulty to recompile autodock on any platform
and Debian has it for all possible (and impossible) architectures already. It even runs on your smartphone.... if it runs Debian or Ubuntu:
http://packages.debian.org/unstable/autodock



We'd need a leader, first. The best would probably be if Scripps would agree to either lead such an effort as an "alpha" variant of its efforts - after all they have done most of the work already. The second best is probably if they offered consulting for the technology and synchronization with their science. The third best would then likely to be to start talking to the Docking@Home http://docking.cis.udel.edu/about/science/advanced.php project and learn more about how they do things, how their integration with wet-lab research looks like, if their clients are technically better or at least open for readily available improvements and/or Open Source also on the client side, if they also use ZINC or would be prepared to do so or by some other means would talk to Scripps - which (knowing no more than the web pages tell (or don't tell) me about docking@home) appeals much more to me in about every aspect.

All that said, the scientific and medical aims are too important to allow our 64bit machines work only left-handed, not even speaking about all the extra that could come. Let's see who else drops a note here and in a month time we should then know more.

Steffen