Open Science Summit 2010 – Thursday Review

1 4 Loading

Over the next few days, Singularity Hub will be bringing exclusive coverage of the Open Science Summit 2010 taking place in Berkeley, CA this weekend. The conference is bringing together a wide range of speakers to talk about what 21st century science should look like, with special attention to crowd-sourced, decentralized innovation and development… basically a combination of scientific research and open source.

Topics include synthetic biology, open data access, gene patenting, DIY bio, the future of publishing & peer review, and open source drug discovery. We’ll be highlighting some exciting companies and projects, interviewing a few of the speakers, and generally reporting on presentations that caught our attention. Feel free to tune in at Fora.tv, which is streaming the conference.

So what is open science? There isn’t a clear consensus – and that’s a good thing. The Summit kicked off Thursday with a panel discussion that tackled some of the central questions of the conference. What if all the products of scientific research – data, publications, tools and techniques, you name it – were available to everyone, to use however they wanted, without any restrictions? This radically free and transparent version of research is, to many, exactly where science should be heading.

The two topics that dominated panel discussion were data sharing and open access to publications. Advocates of open access (e.g. pretty much everyone at the conference) argue that sharing data — all data, even ‘failed’ data — would speed up progress and discovery in unprecedented ways. It would also aid transparency in how data is collected and presented, creating a self-policing environment (and maybe avoid future Climategates). But wouldn’t open data access lead to scientists having their hard work scooped? Panelists generally agreed that publicly presenting data ensures the authors’ credit and prevents intellectual theft.

Opening panel, from left: Victoria Stodden, Peter Hoyt, Cameron Neylon, and Peter Murray-Rust

Some scientists argue that making data available prior to publishing it in scholarly journals (e.g. posting preliminary results online) defeats the purpose of peer-review. Some of the discussion felt these fears are unfounded, and that peer review would still retain its current form at publication; other panelists seemed to hint at a more radically crowd-sourced overhaul of scientific publishing, retaining the standards of publication without its traditional institutions. The panel featured Cameron Neylon, Victoria Stodden, Peter Murray-Rust, and Jason Hoyt.

Stanford post-doc Victoria Stodden argued that open science isn’t a radically new idea, but rather a return to the basic principles of the scientific method. Back in 1665, Robert Boyle established repeatability as a pillar of the emerging European science: scientists should detail how they arrived at their results, so that others can independently reproduce those results. This is what the “Methods” section of published papers is all about. But roughly 20 years ago, the advent of computational data analysis has ushered in a privacy to scientific methods, which nowadays often rely on proprietary computer code. Stodden argues that open science is necessarily also open source regarding its code. Not only would code sharing restore repeatability to the method, it would make computational tools available to other researchers who could benefit from their application.

A few different speakers discussed the future of online databases.  Jason Hoyt gave an excellent overview of Mendeley Research Networks, a database where users curate, cross-link, and discover major journal articles (think PubMed + Digg + a bibliography builder). The site boasts an impressive 450,000 users and over 29 million papers, and it’s still only in beta. Martha Bagnell showed off The Third Reviewer, a site for anonymous (i.e. honest) feedback on publications in both neuroscience and molecular biology. Finally, DJ Strouse and Casey Stark launched their site CoLab, which works like a collaborative wiki for scientists and researchers to solve problems together (and integrates lots of social media concepts along the way).

Mike Gretes discusses the health gap in disease research

Other highlights: Mike Gretes of Mind the Health Gap gave a nice talk on life expectancy differences between developed and developing countries, emphasizing the need for better funding into malaria, HIV, and tropical disease. Peter Murray-Rust pushed open source, open theses, open bibliographies, open citation, open data access… basically open everything (he works – surprise! – on the Open Knowledge Foundation). And Morgan Langille talked about BioTorrents, a protocol for peer-to-peer file sharing of research data.

Aaron will be covering some of the DIY bio stuff, so I’ll leave that to him.  Stay tuned for more Open Science.

Singularity Hub is a media sponsor of the Open Science Summit.

Discussion — 4 Responses

  • Jeff Shrager July 31, 2010 on 8:19 pm

    I was not at the summit, but watched nearly the entire thing on Fora.Tv (still watching). Cred, Creed, and Comment:

    Cred: I am a computer scientist, computational biologist, and psychologist of science. I do science, study science, and publish science and on science in many venues. I have published over 70 peer reviewed papers in all of these fields. Moreover, I am the inventor and PI of the world’s earliest and probably still best cloud-based integrated bioinformatics platform (see: http://www.biobike.org) — and, yes, it’s open source, of course! Look me up — I’m sure you will.

    Creed: I strongly believe in open science in its largest sense. Many of my publications are about open science, and many are in open journals. (I recognize that open journals is not all of open science.)

    Okay, enough for cred and creed, on to …

    Comment: I have high expectations for Open Science, but I had low expectations for this summit. Unfortunately, my expectations greatly exceeded reality. Yes, I wrote that correctly; If this meeting represents the the state of the field of open science, I’m afraid that I have no better word for it than: “Ugh!” What a bunch of ignorant fanboys (and girls); there were so many stupid claims about what makes science work or not work, and so many dumb ideas about what to do about it, that I found myself yelling at fora.tv at least once during almost every talk (sometimes out loud, which startled the folks in my office). I was also watching the fora.tv chat and backchan.nl discussion, which was even lamer than the talks. The CoLab thing that you called the “high point” was pretty much a plain vanilla forum. (With voting and all! Please spare us!) Later that afternoon Michael Nielsen gave a talk which started out by listing the pile of similar things that had failed. Fail. Fail. Fail. Fail. And Fail.

    (I thought that Neilsen was going someplace sensible — to the point that successful science is a management problem, not a technical problem — but his contribution turned out to be equally simplistic: credit preprints….uh, okay, but all that does is to push the politics of publication into preprints, as demonstrated by the fact that the arXiv preprint repository isn’t actually open, but has its own set of rules of contribution and credit assignment (unmentioned by Nielsen). Next idea!)

    Let me say again out of the parens: Making science work is a MANAGEMENT problem, NOT a technical problem. There is, of course, technology involved, but technology is only an enabler. Regardless of how efficiently you can dump your crap online and pull it down, and regardless of how many people you can talk to at the same time, what one needs to make science — open or otherwise — go is leadership and management (and, of course, money helps). BTW, this is the case for ANY project, large or small. Really small projects can be lead and managed by sole investigators, but just because amateur astronomers can spot comets in our solar system (as one person gave at some point as a dumb analogy to open science) doesn’t mean that everyone on earth with a 3 inch telescope will be able to do what the Hubble does! No one’s 3 inch telescope can do that. (And no, you can’t just wire them all together to make a giant distributed telescope bigger and better than the Hubble — Nice try! I had that obvious idea too, and asked a leading astronomer who said to forget it for some reason that I didn’t completely understand — something about the impossibility of accounting for distributed interference.)

    This dumb remark came up in a session on DIY Biotech, which is sort of fun, but has nothing to do with open science. (All that stuff has been “open” for years.) The “open” part of this (that makes it appropriate for an open science summit.) seems like its mostly that folks are building their own PCR machines and gels, for Christmas sake, and are — wonder of wonders — sharing the plans for how to build one in your kitchen … On The Internet (of course)! Sharing the plans for a technology that’s been well understood for 20+ years — Wow!

    Then there were some random cybersecurity folks trying to tell us not to worry about the DIY Biologists. Why? (a) It’s hard to make anything dangerous. (I’m a molecular biologist, and I can tell you that it’s not.) And (b) because someone out there — we being the eyes and ears of Big Brother — will notice it and tell the feds before it’s too late. Uh, right. Thank you for these deeply moving insights. These folks were not at all reassuming regarding who is protecting us!

    And today there are some people flaming about how they are going to discover their own drugs because the pharmas won’t because they won’t make their money back on it. (Ever wonder whether there’s actually a good reason for this — if you don’t think that drug discovery is hard, think again!) [Full disclosure, I not only work in that field, but I actually believe this argument; I just don’t think that the speakers did any justice at all to the complexity of the project.]

    As I said above, I have high hopes for Open Science, but wasn’t hoping for much from this summit, and got far less than I had hoped for from it. Mostly it seemed like a love fest for twitter fanboys (and girls) who think that everything’s a Web 2.0 problem. My theory: If you go into this field with that attitude, you can expect: Fail. Fail. Fail. Fail. And Fail.I hope I’m wrong because so far it looks like that’s where things are going.

    • Jeremy Jeff Shrager August 1, 2010 on 7:27 pm

      “(And no, you can’t just wire them all together to make a giant distributed telescope bigger and better than the Hubble — Nice try! I had that obvious idea too, and asked a leading astronomer who said to forget it for some reason that I didn’t completely understand — something about the impossibility of accounting for distributed interference.)”

      http://medusa.as.arizona.edu/lbto/AO/AOpressrelease.htm

      Now, correcting atmospheric disturbance is definitely not at point of being applicable in your purposed 3-inch distributed telescope array, but to claim its impossible is ridiculous. It wasn’t a question to ask an astronomer anyways, it was a question you should of asked an engineer. Scientists usually only have a working knowledge of the devices they use and their pitfalls, considering the devices they use often require teams of PhDs to design.

      “The “open” part of this (that makes it appropriate for an open science summit.) seems like its mostly that folks are building their own PCR machines and gels, for Christmas sake, and are — wonder of wonders — sharing the plans for how to build one in your kitchen … On The Internet (of course)! Sharing the plans for a technology that’s been well understood for 20+ years — Wow!”

      Taking understood software and slashing the price down to something affordable by a home user is what open source has always been best at – how is that not a useful thing? In fact, open source -innovation- is something relatively new.

  • Jeff Shrager July 31, 2010 on 4:19 pm

    I was not at the summit, but watched nearly the entire thing on Fora.Tv (still watching). Cred, Creed, and Comment:

    Cred: I am a computer scientist, computational biologist, and psychologist of science. I do science, study science, and publish science and on science in many venues. I have published over 70 peer reviewed papers in all of these fields. Moreover, I am the inventor and PI of the world’s earliest and probably still best cloud-based integrated bioinformatics platform (see: http://www.biobike.org) — and, yes, it’s open source, of course! Look me up — I’m sure you will.

    Creed: I strongly believe in open science in its largest sense. Many of my publications are about open science, and many are in open journals. (I recognize that open journals is not all of open science.)

    Okay, enough for cred and creed, on to …

    Comment: I have high expectations for Open Science, but I had low expectations for this summit. Unfortunately, my expectations greatly exceeded reality. Yes, I wrote that correctly; If this meeting represents the the state of the field of open science, I’m afraid that I have no better word for it than: “Ugh!” What a bunch of ignorant fanboys (and girls); there were so many stupid claims about what makes science work or not work, and so many dumb ideas about what to do about it, that I found myself yelling at fora.tv at least once during almost every talk (sometimes out loud, which startled the folks in my office). I was also watching the fora.tv chat and backchan.nl discussion, which was even lamer than the talks. The CoLab thing that you called the “high point” was pretty much a plain vanilla forum. (With voting and all! Please spare us!) Later that afternoon Michael Nielsen gave a talk which started out by listing the pile of similar things that had failed. Fail. Fail. Fail. Fail. And Fail.

    (I thought that Neilsen was going someplace sensible — to the point that successful science is a management problem, not a technical problem — but his contribution turned out to be equally simplistic: credit preprints….uh, okay, but all that does is to push the politics of publication into preprints, as demonstrated by the fact that the arXiv preprint repository isn’t actually open, but has its own set of rules of contribution and credit assignment (unmentioned by Nielsen). Next idea!)

    Let me say again out of the parens: Making science work is a MANAGEMENT problem, NOT a technical problem. There is, of course, technology involved, but technology is only an enabler. Regardless of how efficiently you can dump your crap online and pull it down, and regardless of how many people you can talk to at the same time, what one needs to make science — open or otherwise — go is leadership and management (and, of course, money helps). BTW, this is the case for ANY project, large or small. Really small projects can be lead and managed by sole investigators, but just because amateur astronomers can spot comets in our solar system (as one person gave at some point as a dumb analogy to open science) doesn’t mean that everyone on earth with a 3 inch telescope will be able to do what the Hubble does! No one’s 3 inch telescope can do that. (And no, you can’t just wire them all together to make a giant distributed telescope bigger and better than the Hubble — Nice try! I had that obvious idea too, and asked a leading astronomer who said to forget it for some reason that I didn’t completely understand — something about the impossibility of accounting for distributed interference.)

    This dumb remark came up in a session on DIY Biotech, which is sort of fun, but has nothing to do with open science. (All that stuff has been “open” for years.) The “open” part of this (that makes it appropriate for an open science summit.) seems like its mostly that folks are building their own PCR machines and gels, for Christmas sake, and are — wonder of wonders — sharing the plans for how to build one in your kitchen … On The Internet (of course)! Sharing the plans for a technology that’s been well understood for 20+ years — Wow!

    Then there were some random cybersecurity folks trying to tell us not to worry about the DIY Biologists. Why? (a) It’s hard to make anything dangerous. (I’m a molecular biologist, and I can tell you that it’s not.) And (b) because someone out there — we being the eyes and ears of Big Brother — will notice it and tell the feds before it’s too late. Uh, right. Thank you for these deeply moving insights. These folks were not at all reassuming regarding who is protecting us!

    And today there are some people flaming about how they are going to discover their own drugs because the pharmas won’t because they won’t make their money back on it. (Ever wonder whether there’s actually a good reason for this — if you don’t think that drug discovery is hard, think again!) [Full disclosure, I not only work in that field, but I actually believe this argument; I just don’t think that the speakers did any justice at all to the complexity of the project.]

    As I said above, I have high hopes for Open Science, but wasn’t hoping for much from this summit, and got far less than I had hoped for from it. Mostly it seemed like a love fest for twitter fanboys (and girls) who think that everything’s a Web 2.0 problem. My theory: If you go into this field with that attitude, you can expect: Fail. Fail. Fail. Fail. And Fail.I hope I’m wrong because so far it looks like that’s where things are going.

    • Jeremy Jeff Shrager August 1, 2010 on 3:27 pm

      “(And no, you can’t just wire them all together to make a giant distributed telescope bigger and better than the Hubble — Nice try! I had that obvious idea too, and asked a leading astronomer who said to forget it for some reason that I didn’t completely understand — something about the impossibility of accounting for distributed interference.)”

      http://medusa.as.arizona.edu/lbto/AO/AOpressrelease.htm

      Now, correcting atmospheric disturbance is definitely not at point of being applicable in your purposed 3-inch distributed telescope array, but to claim its impossible is ridiculous. It wasn’t a question to ask an astronomer anyways, it was a question you should of asked an engineer. Scientists usually only have a working knowledge of the devices they use and their pitfalls, considering the devices they use often require teams of PhDs to design.

      “The “open” part of this (that makes it appropriate for an open science summit.) seems like its mostly that folks are building their own PCR machines and gels, for Christmas sake, and are — wonder of wonders — sharing the plans for how to build one in your kitchen … On The Internet (of course)! Sharing the plans for a technology that’s been well understood for 20+ years — Wow!”

      Taking understood software and slashing the price down to something affordable by a home user is what open source has always been best at – how is that not a useful thing? In fact, open source -innovation- is something relatively new.