Monday, July 16, 2007

Rated my first paper in PLoS One

Well, PLoS One has announced its rating system is open. See Chris Surridge's Blog for more detail here. This is basically the first good system for trying out ratings for scientific publications in much the way it is done at YouTube and other Web 2.0 type systems.

I personally am not sure if I think PLoS One is the perfect system for scientific publishing. But we desperately need to try out new systems and as an Open Access Web 2.0 scientific publishing system, the concept is worth trying. So - I decided to try it.

I rated my first paper. I picked a paper by Xiang Xia Min and Donal Hickey on DNA barcoding. Why did I pick this paper. Well a few reasons. First, when I searched PLoS One for "metagenomics" and "metagenomic" it was one of the papers that came up. As soon as I saw the Barcoding title line I knew I had to look at this paper. Why? Well a few years ago, I posted a "dissent" on Faculty of 1000 about a paper on Barcoding. I stand by my dissent in that article. Basically I said that Barcoding sounded a heck of a lot like what Pace and others have been doing with rRNA surveying of uncultured microbes for a long long time. And it seems to me that the barcoding researchers are rediscovering some of the same findings and challenges that Pace and others found. Of course, not all can see my Faculty of 1000 review. Why? Well it is not Open Access. Bummer. Especially since the response by Hebert was unconvincing too ... he basically said that rRNA was not used to delineate species. Apparently he had never heard of phylotypes or any of the extensive work using rRNA to do EXACTLY what they are suggesting mitochondrial genes could be used for in plants and animals.

Anyway - enough about Faculty of 1000. In the new paper, the authors discuss some aspects of barcoding and in particular they address whether one gene can represent all the genes in a mitochondrial genome. Basically, they say yes. But that is beside the point here. You want to know more. Read there paper here. And my comment here. That is yet another benefit of Open Access.

I call on everyone out there to find a paper in which you are interested and rate it and comment on it at PLoS One.

Min, X., & Hickey, D. (2007). DNA Barcodes Provide a Quick Preview of Mitochondrial Genome Composition PLoS ONE, 2 (3) DOI: 10.1371/journal.pone.0000325


  1. As I was reading your post, I was getting ready to comment on how F1000 is not open access . . . and then you went and mentioned it yourself!

    I agree with you that of course individual mt genes will be consistent with each other -- they're all on a molecule with, essentially, no recombination. What's more important is whether mtDNA genes are consistent with nuclear genes and whether individual nuclear genes are consistent with each other. Doing any evolutionary analysis (whether it be phylogenetic or popgen) with a single marker is not advisable -- and an entire mt genome is still just a single marker.

  2. 1) DNA barcoding is not phylogenetics.

    2) I fail to see why it is relevant whether someone has already been doing a similar thing in a totally different context. DNA barcoding is meant to allow the recovery of existing formal species names (which don't exist in the rRNA system you're talking about for the most part), and to delineate clusters potentially in need of formal description. Of course the idea is not entirely new to use DNA sequences for identification -- but a standardized reference system for all animals sure is.

  3. Well, I think it is relevant to understand barcoding to compare it to rRNA diversity studies where people are using a single gene to attempt to identify, count and classify organisms. As for delineate clusters in need of a formal description that is basically one of the main points of rRNA PCR based identification of OTUs (operational taxonomic units - also known as phylotypes). Furthermore this is directly relevant to metagenomics because one of the main goals of metagenomics is to connect those same rRNA clusters to the biology of the organisms, possible through genome sequencing.

    As for a standardized reference system for all animals - that would be nice but again there are parallels to rRNA here as when you publish a new description of a bacterial or archaeal species there one is in essence forced to show it is a novel rRNA type.

  4. While it's noble that a few folks are starting to promote an open, continuous discussion of the published literature. I think this PLoS One experiment is doomed to failure. To really be useful, a large number of articles need a large number of comments and that just isn't happening until were forced to do it. I don't think most editors have folks knocking down their doors asking to review papers; no, they have to seek out reviewers; and then they have to remind them and remind them to turn the reviews in on time.

    I think we need a new system where comment and discussion are integrated and part of the publication process rather than optional add ons.

  5. J

    I get your point and it is a valid one --- the question of how we get people to do peer review is difficult. Everyone is really busy and this is just one more thing. So we need a better reward system for reviews - sure that might help.

    But I am not so sure that the limitations are as large as you seem to think. The new generation of students seems eager to do for science what they do in all other aspects of their life with blogs and Web 2.0, etc. I think it is likely they will be very open about peer review and about participating in things like PLoS One. Sure - it would be good to have more carrots. But I am not convinced that PLoS One have the problems you suggest. It is pretty new. Give it time. After all, it took time for Amazon and other ratings systems to get used and now they are a key part of online sales.

  6. I don't think we need a reward system. I just think that if you plan submit a paper to a journal, the journal should require all of the authors to review or comment on a paper first, so that there is some balance in the system.

    Amazon is playing a different kinda game probabilistically. For example, let's assume "Harry Potter and the Deathly Hallows" sold 15 million copies (an underestimate for sure). Amazon has 2705 reviews for this book, so less than 1 in 5000 of those buying the book reviewed the book. At such a rate, I'd guess most PLoS One papers will never be read by enough people to accumulate 4-10 decent comments.

    But I hope your right. Hopefully I am just one of the doubters in the new generation of blogging web 2.0 students.