To Carole Goble (myExperiment): did you experience network effects? Carole says, a bit of seeding and curation needed at the beginning to put a foundation in place and ensure users of quality. Since trust is so important, a process for creating a safe environment is crucial, a place where people still retain some control over how much they share, etc. - Shirley Wu

Getting users to tag things can still be a challenge. Might be because tagging was in an unnatural place in the workflow. - Shirley Wu

Sean Mooney from Indiana University: "If you build it, they WON'T come." Note, he helped develop Laboratree, one of the other so-called "facebook for scientists". He says one thing they've emphasized is that funding institutions, grant providers etc appreciate when researchers indicate they will deposit, share, make available their data, or use these collaborative tools - Shirley Wu

Nigam: something to talk about is the contributor to user ratio. In wikipedia, it's really low, millions of users to thousands of contributors. In science wikis, the ratio is closer to 1. You can't crowdsource effectively with a ratio like that - Shirley Wu

Drew Endy now talking about OpenWetWare, a wiki resource that started in his lab at MIT by students. When his students started sharing protocols and resources on a lab wiki, other labs took notice and it's now pretty big, ~500 active users. But now the funding has stalled out and they want to get some data on how much impact OWW is having on research - Shirley Wu

Showing a slide: Idea cycle --> research cycle --> publication cycle. Where are the gaps? what are the priorities? Open science not just open data, but all three cycles - Shirley Wu

Example: current publishing cycle way too slow compared to current state of collaboration and authoring tools like Google Docs. - Shirley Wu

Phil Bourne: there's a dichotomy. Do we just refactor things we already have, or do things completely differently? Refactoring can be inefficient but it can be effective in the short term - Shirley Wu

Steve Brenner asks Drew Endy: what metrics could you use to measure usable outcomes for OWW? Drew answers: one metric for him could be has any user recorded all of a research progress on the wiki? How long did it take? etc. - Shirley Wu

The problems of lack of concrete success stories is not only true for OWW but much more general for open science. It looks a bit like a chicken and egg problem of lack of identifiers/rewards that create incentives and lack of success stories to justify the changes in rewards. - Pedro Beltrao

Heather Piwowar mentions: Information behavior issues like these are of great interest to information science field - lots of research and conferences going on so useful place to look for ideas - Shirley Wu

Cameron mentions one example of completely open recording of the complete research cycle: Jean-Claude Bradley and UsefulChem - Shirley Wu

Citation and credit is tricky - a lot of people might be using it but don't think to cite it. "No one cites infrastructure." "You know you've been successful when you don't get cited." Also not always clear what it is you're supposed to cite with all these new-fangled web objects - Shirley Wu

Phil Bourne: idea of "tokens" issued for contributions. But lots of grey areas. What are the relative merits of reviewing a grant vs a paper vs writing a blog post etc. Also what makes sense in one field doesn't necessarily translate into other fields. - Shirley Wu

One thing we could do to make science more open? Carole Goble - change whole idea of citation, and ensure persistence of identity, sustainability. Phil Bourne - needs to come from funding, funders need to be pro-active. Enormous strides already made in Open Access. - Shirley Wu

Show of hands from those with refereed grants - how many have made comments int heir grants about data sharing and availability? Most PIs in the room raised hands. Larry Hunter mentions that many grants don't get funded if you don't mention data sharing or how you will make your research and results accessible. But there are cultural differences - genome field shares more than cancer field, e.g. - Shirley Wu

Russ Altman notes that many studies especially clinical ones, it's not a matter of not wanting to share but it's often a lifetime of work following a very specific clinical cohort, and their entire career depends on them publishing 10-20 papers on that cohort etc so there are other considerations. Cameron says but what if instead of 10 or 20 papers they could get 40 papers out of collaborations? But that is open for debate - Shirley Wu

Nigam Shah: One thing we could do to make science more open? Demonstrated utility and return on investment - Shirley Wu

Heather Piwowar: One thing we could do to make science more open? Be brave. Be brave in being the change you want to see. - Shirley Wu

Larry Hunter quotes ____ McClure: "New science out of other people's data" - Shirley Wu

Nigam notes: openness is meaningless without context and annotation. If you don't know what the parameters are for the experiment and the data, the data is useless - Shirley Wu

Drew Endy notes: surprising to him that he has never come across a community of people whose job it is is to make research better. Mike Wong from SFSU mentions there are some people who provide infrastructure for scientists. But problem is that previously not permanent staff, just transient students etc - Shirley Wu

Drew Endy: Real need to combine the social support that is often there with real technological development. Audience member: problem is that these activities (developing and improving research infrastructures) not often recognized as research. - Shirley Wu

The success of OWW comes the community, not the technology - Graham Steel

So now Drew Endy's One Thing We Could Do: incentivize infrastructure R&D - Shirley Wu

Phil Bourne: we may need to combine a top-down (funders, policy-makers) with a bottom-up (scientists, grassroots) approach and meet in the middle - Shirley Wu

Dave de Roure: examples of this in the UK with funded "Virtual Science Environments" - Shirley Wu

Phil Bourne: key component is to demonstrate that we have impacted science in specific ways. - Shirley Wu

Dave de Roure's One Thing We Could Do: Connect and present success stories - Shirley Wu

Larry Hunter notes: the National Centers for Biocomputing were conceived partly to accelerate scientific discovery through infrastructure and tools. 1. Driving biological problem. 2. Develop tools to solve problems. 3. Demonstrate impact of tools. - Shirley Wu

Quo's One Thing We Could Do: Separate the camps. Open vs. non-open. We also need to reflect on our own identity and what makes a scientist. Do we define ourselves in terms of publications? citations? the data? How might this change? How do we take this into account? - Shirley Wu

It occurs to me that a lot of people in this discussion are established PIs. Just this fact is a huge difference from some of previous discussions I have seen in real life or online where we would discuss the need to get PIs to be aware of these topics. - Pedro Beltrao

Carole: mentions a big fear which is mis-representation. What if someone uses your open protocol for a completely inappropriate purpose and then cites you? These are important issues to address - Shirley Wu

thank you, cameron! took the words right out of my mouth ... - Kaitlin Thaney

Have only caught bits of this but nice wrap up Cameron - Graham Steel

Would have loved to be there - Deepak Singh

Ditto Deepak - looks like a very productive meeting - Jean-Claude Bradley