While I don't necessarily agree with all your points, I really like your attitude. - Pawel Szczesny
Feel free to lay your perspective out Pawel! I'm all ears. There's nothing like discussion to sharpen one's understanding of one's position. - NatBlair
I like it too, and I think you can mostly ignore the "tellin' it like it is" bluster physioprof adopts. - Mr. Gunn
Will be trying to liveblog over the day here. Will see how it works out. - Cameron Neylon
BBSRC - can give steer from funder perspective but wanting to see community views and also educate where necessary. Interested in understanding sociological perspectives so bringing together the biologists and sociologists in this meeting. What are the key issues for BBSRC in data sharing? - Cameron Neylon
ESRC genomics network - research into sociology of data sharing in the biosciences. Issues of scale, organisation, technology. Social sciences trying to catch up with what is happening - Cameron Neylon
Been a request for me not to use names so I won't. Just report themes and comments. - Cameron Neylon
Research Information Network is starting up a project on Web2 tools in science - how are they useful? - Cameron Neylon
Mix of funders, scientists, sociologists and information managers - interesting group they've put together certainly - Cameron Neylon
Need both the willingness to share data and the tools to do so. NCRI working on systems for linking cancer resources together. Also interested in facilitating cultural change. Also talking to pharma about releasing data and trying to encourage that. Which company was it that made the recent announcement on pre-competitive data being made available? - Cameron Neylon
Is it possible to solve social issues with a technical solution? How to ensure acknowledgement for tool developers. Understanding the social issues associated with ontologies - move from OWL to SCOS(?) - do the possible technical solutions degrade quality? - Cameron Neylon
Idea of putting data sharing statements from grants in the public domain so people can check. How does BBSRC monitor compliance with data sharing. Issues of not very good data - should it be made public. Again we come back to the issue of sharing data that is not peer reviewed...and administrative load and predictability of that load, possible issues with IP that may arise in retrospect. The issue of 'hostile mining' of data e,g, animal experiments, stem cells etc. - Cameron Neylon
Answer back which is what I would have said - its important its out there and then you need a way to assess the quality. Peer review does not gauruntee data quality - Cameron Neylon
Science ethics course described given by scientists, lawyers, ethicists, economists etc - our (scientist's) perspective is just a perspective, and may not be a very good one. Other disciplines have very different (and possibly valid) perspectives. Potential ethical issues in pursuing sharing policies. - Cameron Neylon
Hit squad at NERC which aimed to provide every possible assistance to labs to get data into a database. Would go out, install computers, provide tools, support, be friendly (not sure about making coffee) etc etc. Sounds wonderful! - Cameron Neylon
Comment that the obsession with access control is less in companies (anyone in the company is perceived to have a 'right to see') than it is in academia. Some experiences that access control is more important to people than the tools. Scientists want control of where, how, and who can access their data. Prior disclosure is a serious issues for people coming to data sharing. - Cameron Neylon
Dataverse project- a loose database that can provide the data in reusable form close to the published papers. Interest from sociologists and cancer researchers (came out of WGS and gene expression data) - Cameron Neylon
JISC project to look at economic benefits of sharing data. Work has been done on costs but relatively little on economic benefits. - Cameron Neylon
Publications can't cope with presenting the data in their current form. If a data set is a terabyte, what goes in the paper? Current data transmission is insufficient. Funders not getting value for money. - Cameron Neylon
This was an introductory session with everyone introducing themselves - hence the collection of fairly scattering points - Cameron Neylon
BBSRC background - Data sharing policy: Policy has been in place for just over a year. Complements existing best practice guide for curation of data. 10 year period in a useable form is the BBSRC standard. *Biology is changing, large datasets, increasing use of datamining, increasing focus on QA and metadata. *Data sharing picked out early on as key area for Tools and Resources Strategy Panel, BBSRC should have leadership role in promoting appropriate research culture - Cameron Neylon
* BBSRC does not own data outputs (rights reside with institutions) but as a funder has interest in value for money. Is also a grant applicant (to government) *Development of data sharing policy to be inclusive and researcher led. Took three years over two stage consultation process - Cameron Neylon
From consultation exercise: Data sharing is a good thing! Critical (and well established) in some areas. Data sharing needs very different across different domains. IP and data sharing are tensioned. Should be no central silo. Decentralised sharing is preferred. Data should be close to the scientists (not sure what that actually means in practice - given how this flows into issues with Institutional repositories) - Cameron Neylon
Policy aim has two roles: Enforcement and aspirational signpost. Recognises the cost issues. Point out that data sharing is different to archival particularly with respect to costs. - Cameron Neylon
Other funders take different approaches - e.g. NERC has much more centralised approach. Potentially different issues in different domains. * BBSRC focusedd on big data and long time series but reserve the right to be more prescriptive in specific cases - Cameron Neylon
Where resources exist for sharing they should be used. If there is a need, create something, and ask for funding to do it! In some cases local ad hoc arrangements may be appropriate. * Will meet data sharing costs through FEC * Will provide resources to develop underpinning tools etc - Cameron Neylon
Issues of funding models - can ask for support in grant but will this work - serious problenmss in repositories charging for deposition. Wellcome trust model for OA where they deal with it and pay for it offline (independent from grant). Also a significant issues in training competent people to manage the curation process, where are these people? - Cameron Neylon
Proposals require a data sharing policy which is refereed by external reviewers against their knowledge of best practice. Expected to report at end of project.Now going into monitoring phase. Onus is on peer review of best practice. - Cameron Neylon
Overall good community buy in so far but big disparity in quality of responses. Some consternation. More mixed outside of established data sharing communities. Not much in the way of resource requests. Early days yet! * Challenges - can you measure value for money? Will it effect a sustained cultural change (or just grantsmanship?). Can understanding of best practice consistently applied by referees and panels? Are there barriers we don't know about yet? What unintended effects? - Cameron Neylon
I've also looked for such functionality and ended doing the drawing in Inkscape. I was thinking about hacking TeXtopo for that purpose, but never actually started. - Pawel Szczesny
I don't think that was troll-feeding, it was a very even-handed troll-smacking. Provided Lenski stops there and doesn't get sucked into the troll's game of endless bad-faith questions, I'd call it an almost perfect response. The only thing that could improve it, imo, is if the raw data were already Open. - Bill Hooker
Not uncommon actually... many old school chemoinformatics tools were developed/used like this. Oh, sure you may use this software in a *joint* project... - Egon Willighagen
I've seen that couple of times, but I thought such license was banned few years ago... - Pawel Szczesny
I would have thought that that would violate the authorship guidelines of most journals. Now if someone makes an intellectual contribution to the preparation of data or its analysis that is different and it may be a fine line but as it's (apparently) worded that is severely dodgy - Cameron Neylon
As far as I know how it works (I heard it from different people who agreed on such license), authorship guidelines are not violated, because in practice you are _forced_ to collaborate on your projects with software authors. Of course collaboration often means that you get help at a level which is usually found in INSTALL and README files, but still it's counted as such... - Pawel Szczesny
We have a similar issue at large facilities with the question of whether the instrument (or other support) scientists should be an author on papers. To me it depends what they contributed. But forced collaboration sounds like a dodgy model to me - pushes you into building tools that _don't_ work properly. - Cameron Neylon
Well, might not be there, but they do have video of one of the talks I did want to watch - Deepak
I'm watching that video now, I really enjoy Wolski's talks. I wonder what will be the impact of this work in the distributed computing research community? - Cornelius Toole
"...beyond everything else, academia has to start giving credit to researchers performing peer-review and make its results publicly available..." Amen to that. - Bill Hooker
I agree in principle Bill but I worry that it will degrade from a meaningful appraisal to bean counting of how many comments you posted or voted on if we try to measure it - Jean-Claude Bradley
What we're doing here right now is evaluation. How to measure it will be determined at some point, but let's not forget that *doing this* is the important thing, whatever metrics are to come. - Mr. Gunn
Agreeing with Mr. Gunn. Now its impossible to predict how those evaluation mechanisms will work (we are probably lacking some components of the system yet). But some sort of automation seems unavoidable. - Yaroslav
This is a very good point - to summarise 'Are the Web(2) sceptics [in higher education] assessing the risk that they will be left behind by customers who expect their education to be provided through Web(2) style routes' - I hadn't really thought through the consequences and rationale of why undergrads are abandoning institutional email addresses but looking through my address book the evidence is pretty startling. - Cameron Neylon
Is the LIMS idea yours? I like that one. I think a killer app in that arena could make a huge difference to science, because once you get a majority of labrats using an electronic notebook you're halfway to getting them to do real science online. Think of it as a gateway drug for Open Science. - Bill Hooker
Left a comment on the LIMS system bit - our blog is going into major re-write over the summer and if we could get it talking programmatically to other services I think it could handle a lot of what is suggested there. Equally if people have views that it needs to change in a specific way (or you want to get in to have a look so that you can see what needs to change) then now would be a great time to tell us - also added a recent changes section to the front biogang page - Cameron Neylon
I wrote a Joomla-based LIMS once: http://groups.google.com/group.... We use it in our lab. The code is terrible, it's unmaintained and unloved and I'd really like to redo it in a modern web framework, if I had the time. - Neil Saunders
The generic/modular Structural Genomics platform reminded me of a Synaptic Leap subproject that is wanting to do the target selection part: http://www.thesynapticleap.org... ... it seems the idea there is to use the 'traditional' target selection techniques alongside a social target selection scheme that uses collective knowledge (crowdsourcing, with a small, smart crowd I guess). Looks like there is a real need ... - Andrew Perry
We actually have a reasonable LIMS, written in RoR, for the UQ SG project. The problem is that target selection is not built in: it requires that you use a bunch of external scripts then upload a CSV file. What I'd like to see is an SG LIMS that starts with the gene and automates both the selection and construct design as far as possible. We need someone who knows RoR and bioruby or alternatively, Django/GAE and biopython. - Neil Saunders
Where's the URL to the ideas? I'd be interested in seeing them :) - Andreas Matern
At the first DIYbio meeting here in Boston there was a similar project mentioned called Plant Genome Project with the help of http://www.metacarta.com/ Genomes of specific plants where mapped out geographically - Ricardo Vidal
Added a quick comment about the soon-to-be-open LIMS we use for next gen sequencing at Sanger. - Matt Wood
I think the LIMS idea is great, but dare I propose some integration with Outlook? Cut and paste ease with images, word docs, and excel files would be a must have for the aim of getting novice users to start using it. - Mr. Gunn
Nooooooo!!!!1111 Outlook is the spawn of Satan!!!1111 Though I absolutely agree that bringing novices on board will require integration with the tools with which they are already familiar. I just don't think anyone should encourage Outlook use. - Bill Hooker
I have to agree with Bill - with Gmail, Outlook can't compete with speed, convenience and reliability - and it is free to boot. Even our university is switching students over to Gmail because they can't compete with the offerings of storage space. - Jean-Claude Bradley
the Gmail thing is particularly close to me because a month after I switched from Outlook I had a massive crash and was hardly affected - Jean-Claude Bradley
I re-route all mail to GMail. Our uni has 200MB storage and rubbish spam filtering - so why would I bother with them? Wish they were as far-sighted as Drexel. - Neil Saunders
A while ago I hunted around for the Sargasso Sea datasets, hoping to find sequences with GPS coordinates to do a Google Maps mashup. I couldn't find any longitude + latitude data at the time, but maybe I missed it ... does anyone know if that is available now ? - Andrew Perry
@Andrew - don't know about the original Sargasso data, but the Global Ocean Surveyor data at NCBI is annotated with lat/long/depth. Search NCBI all databases for "global ocean surveyor" or here's an example: http://www.ncbi.nlm.nih.gov/en... - Neil Saunders
Neil: Nice ! That might be a project for the weekend then :) - Andrew Perry
I'm seeing a simple search form to the left; a Google map panel to the right; clickable markers that appear to indicate search hits; click marker -> display more fields and links. Mmm yes, could be rather nice. - Neil Saunders
