From biopython at maubp.freeserve.co.uk Fri Jan 29 05:36:40 2010 From: biopython at maubp.freeserve.co.uk (Peter) Date: Fri, 29 Jan 2010 10:36:40 +0000 Subject: [emboss-dev] [Bioperl-l] [MOBY-dev] OpenBio solution challenge: Project updates at BOSC 2010 In-Reply-To: References: <20100128203505.GG40046@sobchak.mgh.harvard.edu> Message-ID: <320fb6e01001290236l1ad02515w403a19f94dbb6d15@mail.gmail.com> Hi all, This is a great topic but should be continue it on just the one mailing list? Is there a suitable BOSC list, or how about the general Open Bio list? On Thu, Jan 28, 2010 at 9:17 PM, Mark Wilkinson wrote: > > Brad, this sounds exciting! > > One thing strikes me, though - by asking for the sub-projects to propose > the "grand challenge" themselves the one thing you can guarantee is that > the "grand challenge" is solvable (or more likely, already solved!) > > Other "grand challenge" kinds of meetings have an independent third party > pose the problem that has to be solved, and then all groups work toward a > solution and compare their results. ?This would, IMO, be more revealing of > the "state of the art" in each Open-Bio project, and point out where the > weaknesses are that we should be focusing on... ?Someone (for example, > you!) could act as the moderator to ensure that the "grand challenge" was > at least a reasonable one, within the scope of what an Open-Bio project > *should* be able to solve... > > Just my CAD $0.02 > > Mark One possible problem with having Brad act as moderator is his ties to Biopython (plus it would be a shame if we'd be one man down for trying to solve the challenges - grin). Having a project representative "sign off" on the challenge might work - or simply the whole of the BOSC committee which is quite balanced. Alternatively some kind of panel of challenges does seem a good way to reduce individual project bias (as suggest by Scooter), but there will still need to be a judging committee. I'm curious what kind of challenges the BOSC committee had in mind - would something like taking a newly sequence bacteria and producing an automated annotation as a GenBank, EMBL, or GFF file be too ambitious for example? There are already several major projects to do this e.g. RAST http://rast.nmpdr.org/ Peter (@Biopython) From biopython at maubp.freeserve.co.uk Fri Jan 29 10:36:40 2010 From: biopython at maubp.freeserve.co.uk (Peter) Date: Fri, 29 Jan 2010 10:36:40 +0000 Subject: [emboss-dev] [Bioperl-l] [MOBY-dev] OpenBio solution challenge: Project updates at BOSC 2010 In-Reply-To: References: <20100128203505.GG40046@sobchak.mgh.harvard.edu> Message-ID: <320fb6e01001290236l1ad02515w403a19f94dbb6d15@mail.gmail.com> Hi all, This is a great topic but should be continue it on just the one mailing list? Is there a suitable BOSC list, or how about the general Open Bio list? On Thu, Jan 28, 2010 at 9:17 PM, Mark Wilkinson wrote: > > Brad, this sounds exciting! > > One thing strikes me, though - by asking for the sub-projects to propose > the "grand challenge" themselves the one thing you can guarantee is that > the "grand challenge" is solvable (or more likely, already solved!) > > Other "grand challenge" kinds of meetings have an independent third party > pose the problem that has to be solved, and then all groups work toward a > solution and compare their results. ?This would, IMO, be more revealing of > the "state of the art" in each Open-Bio project, and point out where the > weaknesses are that we should be focusing on... ?Someone (for example, > you!) could act as the moderator to ensure that the "grand challenge" was > at least a reasonable one, within the scope of what an Open-Bio project > *should* be able to solve... > > Just my CAD $0.02 > > Mark One possible problem with having Brad act as moderator is his ties to Biopython (plus it would be a shame if we'd be one man down for trying to solve the challenges - grin). Having a project representative "sign off" on the challenge might work - or simply the whole of the BOSC committee which is quite balanced. Alternatively some kind of panel of challenges does seem a good way to reduce individual project bias (as suggest by Scooter), but there will still need to be a judging committee. I'm curious what kind of challenges the BOSC committee had in mind - would something like taking a newly sequence bacteria and producing an automated annotation as a GenBank, EMBL, or GFF file be too ambitious for example? There are already several major projects to do this e.g. RAST http://rast.nmpdr.org/ Peter (@Biopython)