Monday, November 27, 2006

Has your scientific research been wasted?

I had a good Thanksgiving weekend this year - spending time with family and friends. But as I go back to work this week I have now gotten somewhat depressed over something I did Sunday night. I decided to remove myself from the UC Davis internet proxy to see how many of my past papers that I have published I can obtain without the UC subscriptions. So I went to pubmed, and typed in my name (Eisen JA) and got most of my papers, which are listed at the bottom of this blog (some do not come up due to publication off the pubmed grid or due to co-authors screwing up my initials). (NOTE  - LISTING DELETED 4/09 BECAUSE THE FORMATTING IS ALL MESSED UP)

And then I went to see how many of my papers were freely available and how many were not. What I was most interested in was - what is the deal with papers I wrote before becoming an Open Access convert? For many it is easy to figure out if they are freely available - Pubmed has a link saying "Free in PMC" which refers to Pubmed Central. For others, it was a little trickier.

The results were both good and bad and a summary is below. A few things struck me. First, a lot of my life's work is not readily available without paying other for it. In the day and age of the internet, this means that these papers will simply be read less and less as time goes by. And that makes me very sad. If I had chosen to publish those papers in other journals, anyone in the world could get them at any time. Thankfully I did publish many papers in journals like PNAS, and ASM journals, and NAR - journals that have now decided to release them to Pubmed Central. And also thankfully (but less so) I published some papers in journals that have at least made them freely available on their web sites.

Most surprisingly to me is that a reasonable number of my papers in Nature are freely available on the Nature web site as part of their Genomics Gateway program. Nature deserves serious kudos for doing this and they stand out compared to Elsevier journals (which do not seem to ever do this) and even Science. This is disappointing as Science is published by a scientific society but apparently does not seem to care much about access to publications. Nature, a commercial publisher, is in my opinion doing more for scientific openness than Science. Now, Nature has a long way to go, but I am SO glad I listened to their editors like Chris Gunter and Tanguy Chouard who made a big deal about the Genome papers being free. I did not think it was that big a deal, but in retrospect they were ahead of me in thinking about availability. Plus Nature clearly makes more of an effort to provide free online material than they have to - and certainly make more available than Science.

So in the end - I am sad about my partially wasted past. But I am pleasantly surprised that at least some papers I thought would be more restricted are actually free (although only on the Publishers site for now - Hopefully these journals will submit them to PMC at some point). I guess - you win some and you lose some and some are somewhere in between.

******************
Summary of openness --- other scientists should do this exercise

In Pubmed Central and Open Access
Available free on publisher's sites (notideal but better than nothing)
Must buy paper
Not available anywhere

17 comments:

  1. While I support, as do you, "top down" open access by the journals themselves, there is also "bottom up" open access by putting PDFs of one's own papers on personal web sites -- these then get indexed by search engines -- for example, the Badger, Eisen, and Ward 2005 paper you list as being closed access may be so on the IJSEM site, but it is on my site and is Google accessible.

    This behavior of posting PDFs is almost universal among computer scientists -- I'm not sure why it seems less common among biologists.

    ReplyDelete
  2. Well, I was kind of purposefully leaving that out of the discussion here. Self archiving is clearly important and I too have posted all my papers on various personal web sites. However, I vehemently disagree with those who think this is the equivalent to the journals diong it themselves. First of all, my web site could easily disappear and then the paper is gone. Pubmed Central is much less likely to disappear. Second, having everything in one database allows for much more thorough analysis and searching and indexing. Yes, you can get the paper through google while it is on your site. But if you want to use textpresso or other new full text search engine methods, they work better in many ways when they do not have to use the entire world wide web as their database. Third, though you and I and physicists and computer scientists and many biologists may use google scholar, or other search engines, many others out there do not. The "availability" of a paper is not just a measure of it being free somewhere. It is also a measure of how likely people are to find it. And that is why getting things in Pubmed is important. This is one of the biggest limitations of self archiving. Since the journals own the copyright on many such papers, they can lobby Pubmed to prevent them from creating links to google scholar and other search engines. Thus that paper in Pubmed only has links to IJSEM and Highwire Press. So - self archiving is a very good thing - but it is not good enough.

    ReplyDelete
  3. Self-archiving works best when done not on personal websites but with OAI-PMH compliant repositories, because then you can search efficiently with something like OAIster and the long-term archiving is taken care of. In a good independent repository, I'd say the work is as safe and available as it would be in PMC.

    The problem I've run into (scroll/search to "personal example") when trying to free up my work is that there is no repository that's a natural fit. This is likely to be the case for anyone in biomed research, unless your own institution has an IR or your work fits into the specific categories of biology accepted into arXiv (quant biol) or Cogprints (behaviour, evolution theory).

    Peter Murray-Rust recently pointed out that KnowledgeForge/CKAN might offer a solution to this problem.

    ReplyDelete
  4. Thanks for the shout-out, dude!

    Actually the policy of Nature Publishing Group is that authors can put their accepted version into PubMedCentral six months after publication. So you could go in and put all of those Nature family papers into PMC, if you still have the author's final version.

    ReplyDelete
  5. How long has this been in place? By what you wrote you imply that it is retroactive for all Nature publications, which would be really cool for me since NOBODY seems to have read my 1997 Nature Medicine article where I first proposed the term "phylogenomics" and gave an example of how evolutionary and genome analysis can be combined to better predict gene functions. The publication does not seem to be online anywhere in the Nature archives.

    Since I am an electronic pack rat. I probably have the final authors versions for almost all of my papers. Any more detail on this policy would be very useful.

    ReplyDelete
  6. Well, I am now not so convinced about the ease of submitting a author's version of a paper to Pubmed Central. I tried to submit a few papers via PMC's suggested submission systems and none seem to work for me due to one key issue. The papers I want to submit were not supported by NIH research and it seems that the tools PMC has to allow author's to submit are only for those trying to compley with NIH or Wellcome Trust rules or guidelines about submitting papers.

    I have only one question for PMC - WHAT ABOUT THE REST OF PAPERS? NIH does fund a lot of work, but not all of it. When I tried to use MyNCBI to submit there is a form requiring you to fill in the NIH grant that supported the work.

    So - Chris - or anyone else out there - do you know how to get this done?

    ReplyDelete
  7. I wrote to you privately as well, but for anyone following this thread: yes the NPG policy is retroactive, and it has been in place for about a year.

    As far as submitting to PMC, unfortunately I have heard it is not always easy, and I would suggest you contact David Lipman or someone at NCBI who has been involved with setting it up. They are also supposed to have a help desk I think.

    ReplyDelete
  8. Stay tuned ... I have a phone call scheduled with Lipman today. The myNCBI system seemed pretty straightforward - at least the parts I was able to get to before it asked for an NIH grant. And I think the NIH tools inside the NIH grants system are not so bad either.

    Some information on NIH grant related submissions can be found at NIH Library Manuscript Submission Assistance

    Perhaps the best info I have found is at The NIH Manuscript Submission System Website

    ReplyDelete
  9. I talked to David Lipman today on the phone. And he was extremely helpful. I guess I did not really think about one issue here - which is that the Welcome Trust and NIH pay the costs of converting authors papers into the right format to submit to Pubmed Central. Thus that is why there is no straightforward way to convert non-NIH non-Welcome funded work.

    Thus I think people should lobby NSF and other funding agencies to step up to the plate provide some support to PMC for converting authors papers funded by their money into PMC papers.

    On a related note Lipman said he will look into options to make it possible for other work to be put into PMC. I will post information on that here if/once I find out.

    ReplyDelete
  10. Interesting post. I would like to take a moment, though, to correct what may be a misperception regarding Science's access control policy -- albeit a misperception that stems from the deficiencies of our current interface, which we're trying to correct.

    Our policy is that all original research papers in the journal since the end of 1996, when Science began putting the full HTML text of research content on the Web, becomes available free of charge 12 months after publication, to all users who have completed a free registration with the site. News and Commentary (such as Perspectives) remain accessible only to subscribers or on a pay-per-article basis.

    Unfortunately, we have not done a good job, in the design of the sign-in page that confronts users who aren't logged in, in making clear the fact that the paper can be accessed without charge if the user registers for free with the site. We are working with our online vendor to try to change that.

    In any event, I'm not sure it's completely accurate to say that you must buy the Science papers you list on this posting; all of the research papers that you list as "must buy" can be accessed free of charge by registering with the site, as can any research paper on the site published more than 12 months ago. This has been our policy since 2001. Unfortunately we haven't made that policy as clear in our user interface as it should be, but we're working to change that.

    Stewart Wills
    Online Editor, Science

    ReplyDelete
  11. Well, you certainly should try and make that clearer because all I see when I am not within the UC Davis system is some page implying I have to pay to get access (maybe it said I have to sign in, but the implication to me was $$$).

    I would say however, that it would be better for openness if a sign in was not required. In addition, does Science allow submission of this material to PMC?

    ReplyDelete
  12. OK, from what I can tell, although it would be useful to have some input from Science folks, Science does not allow for authors to submit their publications to Pubmed Central.

    ReplyDelete
  13. And of course, it would be EASIEST for all if one just published in an open access journal like PloS Journals or BMC journals in the first place

    ReplyDelete
  14. I believe our policy wrt to PubMed Central is as follows:

    "For a paper created under an NIH grant and accepted for publication on or after 2 May 2005, authors may implement posting in PMC, no sooner than six months after final publication, of the 'accepted version' of the paper -- that is, the version of the paper accepted for publication after changes resulting from peer review, but before Science's editing, image quality control, and production. In submitting to PMC, authors of NIH-funded Science papers should set the time of public release of the accepted version at six months after final publication."

    This is all I know about the matter.

    Best wishes,

    Stewart Wills

    ReplyDelete
  15. That policy is pretty disappointing to me since it is biased towards NIH funded research. Why, if Science allows some things to be put in Pubmed Central, would it not allow research funded by other sources like NSF, DOE, etc. I guess from this one can concluse that Science does not support Pubmed Central and only agrees to do this so that NIH funded research can be submitted to Science? It seems pretty lame if that is true. If the article is available for free on Science's web site, why not allow the authors version's to be free in Pubmed Central. That would probably in the long run attract more attention to Science and Science articles. Disappointing to say the least.

    ReplyDelete
  16. Jonathan Eisen wrote "Self archiving is clearly important and I too have posted all my papers on various personal web sites. However... my web site could easily disappear and then the paper is gone. Pubmed Central is much less likely to disappear."

    How about depositing instead in your own Institutional Repository? Your institution is less likely to disappear. And, to be safe, the deposit can then be harvested automatically by PubMed Central and/or other central repositories.

    Stevan Harnad

    ReplyDelete
  17. Many institutions limit access to their online information. Making this information available will be an asset to all.

    ReplyDelete

Most recent post

Talk on Sequencing and Microbes ...

I recently gave a talk where I combined what are normally two distinct topics - the Evolution of DNA Sequencing, and the use of Sequencing t...