Meetings and Seminars
Scientific meetings and seminars are one example were we can do better. Web 2.0 is an ideal approach for this, and Upcoming is the classic application. There are of course a number of websites that list meetings and seminars for scientists, but they either focus on the big meetings or list just the seminars of a particular institution. Look at the discussion How to find a science event in Berlin in the Nature Network Berlin Forum to see what I mean.

What can we do to improve the situation?
We can wait that either one of the big players or a clever startup has a great idea. But one of the attractive features of Web 2.0 is user participation. We need more discussions between scientists and software developers on what is needed and what can be done. These discussions are of course already taking place, but science bloggers can do more to collect interesting ideas and articulate them. We want the integration of reference managers in online writing tools such as Google Docs or Buzzword, but how do we make our voice heard?

Secondly, we can write applications ourselves. The barriers of entry have become really low, and one reason are the APIs (application programming interfaces) of both science applications or conventional Web 2.0 apps:

• Pubmed
• Scopus
• Connotea
• Google Maps
• Google Docs
• YouTube
• Flickr
• Technorati
• Facebook
• eBay
• Amazon
• Yahoo

And there were hints of a Nature Network API. With some skills in PHP, Python, Java or Ruby, anybody could create an interesting mashup with these APIs over a weekend. Maybe linking Connotea tags to YouTube videos and Flickr pictures?

For more information, read this forum post by Matt Brown.

Then I thought about a recent blog post by Gunther Eysenbach: Black sheep among Open Access Journals and Publishers. In this post he calls the sending of unsolicited emails simply spamming and argues that there are also throw-away journals out there from shady publishers trying to cash in on the current surge of interest in open access publishing.

And this is what all three journals mentioned above have in common: they are open access journals and the author pays for the (accepted) article. It is obvious that any journal that gets paid by the author is interested in soliciting articles whereas a subscriber-pays journal would be interested in attracting new readers. There is nothing wrong with this, but there are two potential problems. (1) Like most people I don’t like spam. (2) Journals with an author-pays business model have to be extremely careful about the quality of their papers.

Potential authors should first check whether the journal (if it is a biomedical journal) is indexed in Medline (Genomics Insights is not) and either has a reasonable impact factor or (for new journals) receives enough citations.

Most journals allow posting post-prints on a university website
The copright agreement with the journal is the easy part. Most publishers allow posting of post-prints (after peer-review, but not the journal PDF) in a non-commercial repository, usually the repository of your institution. Below are the policies of three prominent publishers.

Elsevier
As an author, you retain rights for a large number of author uses, including use by your employing institute or company. These rights are retained and permitted without the need to obtain specific permission from Elsevier. These include (...) the right to post a revised personal version of the text of the final article (to reflect changes made in the peer review process) on the author’s personal or institutional web site or server, with a link to the journal home page

Springer
An author may self-archive an author-created version of his/her article on his/her own website. He/she may also deposit this version on his/her institution’s and funder’s (funder-designated) repository at the funder’s request or as a result of a legal obligation, including his/her final version, provided it is not made publicly available until after 12 months of official publication. He/she may not use the publisher’s PDF version which is posted on www.springerlink.com for the purpose of self-archiving or deposit. Furthermore, the author may only post his/her version provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer’s website. The link must be accompanied by the following text: “The original publication is available at www.springerlink.com”.

Nature Publishing Group
When a manuscript is accepted for publication in an NPG journal, authors are encouraged to submit the author’s version of the accepted paper (the unedited manuscript) to PubMedCentral or other appropriate funding body’s archive, for public release six months after publication. In addition, authors are encouraged to archive this version of the manuscript in their institution’s repositories and, if they wish, on their personal websites, also six months after the original publication. In all these cases, authors should cite the publication reference and DOI number on any deposited version, and provide a link from it to the URL of the published article on the journal’s website.

The policies of some of the smaller publishers can be more difficult to find. Sometimes an email exchange with the publisher will be necessary.

Many universities have institutional repositories
OpenDOAR is a directory of open access repositories. My university doesn’t yet have an institutional repository. After a short email exchange they offered to host my post-prints on a public webserver. I currently have no details on the software platform used or how many people in my university use this service.

Post-prints may no longer be available
Most publishers don’t allow posting of the journal PDF. You have to post the final manuscript after per-review (post-print). The problem: I no longer have these manuscripts for papers published more than a few years ago – thinking that the PDF would be enough.

Your post-prints are hard to find
OpenDOAR has a search function, but searching several institutional repositories at once is complicated. Your best bet is probably to find a paper in Pubmed and then try to find the institutional repository for that author. But maybe you have to check several institutional repositories if the authors are not all from the same institution.

Conclusions
Posting your paper in an institutional repository can be a challenging project. It is therefore advisable to think about this before paper submission. What is your publication strategy? Do you need open access? Does the journal offer free content after an embargo period of 6 or 12 months? What is the journal policy regarding post-prints? And most importantly, keep the manuscript version right after peer-review. A central repository such as Pubmed Central doesn’t have most of these shortcomings.

The new ResearcherID features include a mashup with Yahoo Maps that shows the location of your collaborators. Here are some examples from fellow Nature Networkers:

• Martin Fenner
• Raf Aerts
• Bob O’Hara

Wentworthville, Australia is the furthest I can go to find a paper coauthor.

Now the really interesting question would be: when do we see mashups with Nature Networkers? For a start we have the Nature Network London Flickr Mashup created by Matt.

An article and accompanying editorial in this week’s JAMA look at guest authorship and ghostwriting in publications related to rofecoxib (better known as Vioxx, produced by Merck). Guest authorship was defined as an author that does not meet authorship criteria. The Uniform Requirements for Manuscripts to Biomedical Journals from the International Commitee of Medical Journal Editors (ICMJE) define authorship as follows:

Authorship credit should be based on 1) substantial contributions to conception and design, or acquisition of data, or analysis and interpretation of data; 2) drafting the article or revising it critically for important intellectual content; and 3) final approval of the version to be published. Authors should meet conditions 1, 2, and 3.

Ghostwriting was defined as the failure to designate an individual that made significant contributions to the research or writing of a manuscript.

The JAMA article did a systematic analysis of the court documents obtained during litigation related to rofecoxib. Guest authorship was identified in 16% of research articles and 26% of review articles; ghostwriting was identified in 13% of research articles and 10% of review articles. The editorial in the same issue is called Impugning the Integrity of Medical Science: The Adverse Effects of Industry Influence and dicusses this article as well an another article called Reporting Mortality Findings in Trials of Rofecoxib for Alzheimer Disease or Cognitive Impairment in the same issue. The editorial proposes that drastic action is necessary and includes a list of 11 measures.

Criteria for authorship and disclosure of financial interests are clearly defined not only by the ICMJE, but also by the World Association of Medical Editors (WAME) and by most journals. And the European Medical Writers Association (EMWA) has published guidelines on the role of medical writers in peer-reviewed publications. As medical writers usually don’t fulfill the authorship criteria defined above (with the possible exception of review articles), they should rather be acknowledged. The EMWA guidelines propose the following wording:

We thank Dr Jane Doe who provided medical writing services on behalf of XYZ Pharmaceuticals Ltd.

The rules are clear and it is also clear that there will be violations of these rules. Guest authorship and ghostwriting are probably common practices, not only in publications supported by drug companies. Typical examples would be the inclusion of the department head that did little more than to provide financial support or the exclusion of the technician that did critical experiments but is not acknowledged. Cases of suspected guest authorship or ghostwriting should be taken seriously and the papers in JAMA will ignite a renewed discussion on these topics.

For different views on this topic, look at Exit ghost by Juan Carlos Lopez over at Spoonful of Medicine and Cut it out. Cut it out now. by Derek Lowe at In the Pipeline.

Bob O’Hara wonders about the cost of publishing in Open Access: Show us the Money!. He argues that shifting the costs from reader to author can create problems. Most authors, especially those with limited resources, would be reluctant to pay submision fees if they can also submit to a journal without those fees. But the reader-pays model could give authors more bargaining power with journals. The post created an interesting discussion about the different aspects of publishing costs

Graham Steel pointed out the Second European Conference on Scientific Publishing in Biomedicine and Medicine that takes place in Oslo September 4-6 in Oslo, Norway. The conference focusses on open access and bibliometrics.

I wrote two blog posts about public access. In Germany, most research organizations have signed the Berlin Declaration, but in contrast to the new NIH policy, there is no mandatory public access. In another blog entry, I looked at public access to my own research papers – most of them are only accessible for those with institutional journal subscriptions.

What is the next step for me? That I need to learn more about self-archieving – both the policies of the journals I have published and the institutional repository at my university.

• 3 papers available for everybody
• 10 papers only available from within my institution (Journal subscription required)
• 2 papers only available for purchase

Interestingly, the papers in the two journals with the highest impact factor are both available as fulltext. And the third fulltext paper is my paper with the most citations (and published in 1998).

Conclusion: Not that anyone would care what I have to say, but you have to work in an institution with a good library budget to read my papers.

The Berlin Declaration from October 2003 was a strong statement of support for Open Access and was signed by all major research and funding organizations, including Deutsche Forschungsgemeinschaft (DFG), Max Planck Gesellschaft (MPG), Helmholtz-Gesellschaft, Fraunhofer-Gesellschaft and Leibniz-Gemeinschaft. In contrast to the new NIH public access policy (and the Welcome Trust and Howard Hughes Medical Institute), there is no mandatory Open Access in any of these organizations.

The German Publisher Springer, one of the largest STM (Science, Technology and Medicine) publishers, has a Springer Open Choice option. Authors who pay for this option will retain the copyright of their paper and the article will be made available with full Open Access.

The Director of the National Institutes of Health shall require that all investigators funded by the NIH submit or have submitted for them to the National Library of Medicine’s PubMed Central an electronic version of their final, peer-reviewed manuscripts upon acceptance for publication, to be made publicly available no later than 12 months after the official date of publication: Provided, That the NIH shall implement the public access policy in a manner consistent with copyright law.

Starting April 7, the new NIH Public Access Policy implementing this law will take effect for most NIH grantees. The NIH is soliciting comments until May 1st for this new policy. Open Access and this new policy are complicated topics that simply can’t be covered in a single blog post. This week will certainly see a lot of discussion on this topic, both on Nature Network and elsewhere. Please state your view on the topic either in blog entries or join the discussion in the Publishing in the New Millenium Forum. I will try to summarize (some) of the discussion at the end of this Public Access Week.

Although the typical work of a scientist is a very different from a cyclist performing in the Tour de France, a first case of erythropoietin use(or rather misuse) has now been reported in Germany. A neurobiology postdoc working on the role of erythropoietin in Alzheimer disease has apparently had the idea to do a little self-experimenting – probably thinking of the famous experiment by Barry Marashall that won him a Nobel Price. Or he was pictured – during his long experiments lasting well into the night – the famous German cross country skier Johann Mühlegg who tried to win an Olympic Gold medal in Salt Lake City with the help of darbepoetin. What he didn’t think of was the price of erythropoietin. After burning some 20.000 Euro, the technician and then the principal investigator became suspicious. The case is now under investigation by the university and the Deutsche Forschungsgemeinschaft. What the authorities don’t know is how to handle this case. Is this simply misappropriation of the research funds, is this brain doping, or is this a misunderstood brilliant scientist? The World Anti-Brain Doping Authority now got involved to help clear the issues.

1. Pubmed. This topic has been extensively covered in Anna’s post.
2. Bibliographies. The process from finding a paper in Pubmed to storing it into your bibliographic software to downloading and storing the PDF to creating a bibliography in a paper is extremely complicated. In my case this process involves a web browser, Endnote, Papers, a PDF viewer and Microsoft Word.
3. Writing a paper. The idiosyncrasies of Microsoft Word (the tool probably used by most people for paper writing) would make for a different blog post. Microsoft Word can do many things that are never needed in a scientific manuscript, but lacks features that are important. And it likes to crash on long texts.
4. Collective Writing. The process of having several people working on a manuscript and managing different versions is overly complicated.
5. Paper submission. The online tools such as Editorial Manager are nice, but there is still too much formatting work required (in a format that is probably different depending on the journal).

Interestingly, the way we do paper writing and paper submission has changed a lot between ca. 1985 and 1995. Basically the transition from analog to digital. Who still remembers Index Medicus, Current Contents, copying papers in the library, Letraset for figure numbering, mailing hard copies of manuscripts around, etc.?

But what has changed in the last 10-15 years? Probably not that much. Web 2.0 is an overused term, but there is so much potential for improvement in the area of manuscript writing. We should not be satisfied with what Pubmed, Endnote, Microsoft Word, etc. offer today. That is why I don’t understand some of the discussion surrounding Anna’s blog post, basically advising her to just sit down and learn her Pubmed stuff.

If that is not enough for you, some people have already taken the next step: Web 3.0 and health librarians: an introduction.

Wikipedia has become a respectable source of information that rivals the more traditional encylopedias such as the Encyclopaedia Britannica. Remember the December 2005 Nature study that compared the two? Wikipedia has accurate information even on such obscure topics as dwarf woolly mammoths from Wrangel Island.

But we all know that Wikipedia is not perfect. If we want to improve the information on topics we care about – and probably have spent years working on – we could become a Wikipedia editor. Especially if we are interested in an audience that includes not only fellow scientists, but also people from other scientific disciplines, journalists or students.

Writing and editing content on Wikipedia is a complicated process. Because anyone can edit content, a great number of rules exist to make sure that the articles have correct information, don’t violate privacy, aren’t misused as marketing opportunity, etc. All these rules, as well as many practical tips and other tools are of course available online. But Wikipedia: The Missing Manual is a very good text for the aspring new Wikipedia editor that wants a more systematic introduction.

This study is just another confirmation of the six degrees of separation concept. A 2001 PNAS paper found the same for scientific collaboration networks, using common authorship for a paper as connector. The mean distance of two random authors in the MEDLINE database was 4.6.

Based on these findings, you can almost expect the same connections through joint authorship between two random Nature Network members. Now if I could only find a tool that would help with that task. Not everybody is a mathematician and can simply calculate his Erdös number right from the American Mathematical Society website.

• Time. Scientists have little time, and rather spend this time in the laboratory or reading papers
• Trust. Web 2.0 sites for scientists haven’t (yet) build a reputation. For important decisions (e.g. a critical protocol for an experiment), they rather ask a colleague they know.
• Inappropriate Tools. The requirements for scientists are very different from the typical Facebook or Digg user.

I believe that David is right in his analysis that the big majority of scientists (at least in the life sciences) don’t read science blogs or participate in social networking sites like Nature Network. This could mean two things: a) Web 2.0 is not working for biologists or b) we are just at the beginning and need to be patient. As a science blogger I like to believe in b). Fortunately or unfortunately, the success of any Web 2.0 project depends on a large number of users.

As of 2008, I think that Science 2.0 (or whatever you want to call it) has had a good start. But it is very important that we stay focused on where we want to go and not get distracted by the possibilities that the technology offers. One goal I’ve set for myself and have written about in this blog: Web 2.0 should make the process of paper writing much easier. This includes easy access to papers needed for your manuscript (including open access), online writing tools such as Buzzword or Google Docs, online tools for managing or sharing your references (e.g. Connotea or Refworks) and tools for collaboration and coordination (e.g. Basecamp). Right now, the different pieces don’t quite fit together, but the potential is there for tremendous savings of time and money. And that is attractive.

Seven US-educated scientists working at the Max Planck Society’s institutes are facing criminal charges for impersonating a doctor.

This extremely embarrassing story is fortunately no longer possible. The German Kultusministerkonferenz decided on March 6 (text in German) to allow U.S. PhDs to call themselves Dr. in Germany. Until this story evolved, I didn’t even know that this is an issue.

You understand that these are sensitive ethical issues and you want to do the right thing. When is joint first authorship a problem, can I publish a paper a second time if it is written in another language, should my department head be a coauthor, etc. You will of course discuss these questions with your coauthors, but where else can you look for advice?

• Journal author guidelines. The first and most useful place to look for advice, both for technical aspects, but also for ethical issues surrounding your manuscript. Author guidelines probably often are read too late (days before submission) and not carefully enough. The Editorial policies of Nature are found here.

• ICMJE (International Commitee of Medical Journal Editors). The Uniform Requirements for Manuscripts Submitted to Biomedical Journals could be the next place for advice. The text contains a lot of useful information for both authors and editors, including special reporting guidelines such as CONSORT for randomized controlled trials.

• COPE (Committee on Publication Ethics) COPE is another useful resource that I recently discovered. I especially like the Case section, where real-life examples such as ghost authorship or reviewer competing interests are discussed.

If there are still open questions after reading these guidelines, join the regular discussion in the Nature Network blogs and forums on these topics, most recently Nobel prize-winning lab retracts paper from Nature in the Neuroscience forum.

The book is in German and is called Italien – Kurzes Reisehandbuch von Karl Baedeker. The Baedeker series of travel guides (this one is about Italy) is still popular, but was started around 150 years ago. My book is the third edition and was printed in 1895. I will be travelling to the Amalfi Coast near Naples for my Easter holiday and I thought this would be a good opportunity to do some exprimental tourism.

Experimental Travel is an playful new way of traveling that I learned about exactly one year ago on March 6 (another good reason for this post) when I bought this book – in a bookstore in Auckland on my first vacation day in New Zealand. The books describes many travel experiments and laboratory results, including

• Airport tourism. Spend 24 hours in an airport without getting on a plane
• Confluence seeking. Visit ordered points such as exactly 35° S / 117° E using a GPS device
• Mascot travel. Take pictures of your mascot outside famous landmarks, made famous by the French movie Amelie.

The experiment I wanted to do since reading this book was historic travel, i.e. traveling with the help of a guidebook from a different time. Italy is perfect for that, because my 1895 guidebook can still be very useful for many places I will be visiting. Pompei is a good example. I will try to write lab notes and report on the outcome of the experiment.

The problem with posters is that they are second class citizens to oral presentations in most meetings. My personal experience with posters, both my own work and posters from others, has been mixed. In some meetings the poster presentation was a relaxed event (including beer and brezels in the last meeting) with good discussions in front of the poster, but often the poster presentation is not much more than a trick to increase conference attendance.

A recent paper (link to english version) in the German journal Deutsches Ärzteblatt systematically interviewed poster authors and attendees at a German meeting. This meeting used the format of a moderated poster presentation. The attendance in the poster presentations was very low, but was valued by younger scientists and by the moderators. One third of the posters had already been presented at another meeting, an issues that touches the problem of duplicate papers that we recently discussed here on Nature Network. In another study, 12% of the posters had already been published as full paper at the time of the meeting.

In my opinion poster presentations are an important part of every scientific meeting. They should be taken seriously by using a competitive peer review process, including the rejection of abstracts that have already been presented or published. And they should be allowed enough space and time in the meeting schedule. Maybe we could also come up with new formats (e.g. this video in poster) that make the poster presentation both fun and scientifically engaging.

Research findings about cardiovascular side effects of COX-2 inhibitors are at the center of the dispute and Pfizer is now seeking arguments for their case not just from published papers but also in confidential peer reviews and manuscripts that were rejected. This legal dispute is important because it touches central aspects of the peer review process.

The findings of scientific papers can have consequences not only to the scientific community involved, but also for the personal fortunes of their authors (e.g. new jobs or grants), the treatment of patients, our policies towards climate change – or the profit of a drug company. With that much at stake, the temptation to move the scientific argument from the editorial office to the courtroom is there. We should resist this temptation, or the peer review process and the way we communicate science will never be the same again.

Nature Network
You can list your publications in your Nature Network profile. Simply add the DOI or Pubmed ID in the form provided. (Some of) my publications are listed here.

Facebook
Facebook has a similar feature for those unfortunate souls not yet on Nature Network. Use the Facebook application Medline Publications.

Google Scholar
Only Facebook members can see your Medline publications. But everybody can use Google Scholar. My publications are listed here.

Pubmed
Why not use the original? Use the search term “Fenner MH”[Author]. My publications are listed here.

Scopus
Searches in Google Scholar and Pubmed search by name. This means that the links above don’t work if your name is not unique. Scopus uses a unique author ID to overcome these problems. My author ID: 7006600833 and 7006600825. If you have more than one author ID (as in my case), you can ask Scopus to merge them together into one author ID. Scopus also provides an API to create mashups with other data. One example is this mashup with Google Maps that shows a map of the most highly cited papers by subject area. The problem with Scopus? You have to sign up for a user account to use Scopus.

ResearcherID
The new kid on the block. ResearcherID is currently only available to ISI Web of Knowledge users, for all others it is invitation only. But everybody can access your publications. My researcherID is A-7225-2008 and the list of my publications can be found here.

The problem with ResearcherID? ResearcherID is tightly integrated with other Thomson Scientific products. You need a ISI Web of Knowledge account to add papers, Or you can import your citations from a RIS file, a file format from the Thomson Scientific Reference Manager.

As you can see, all these tools have their shortcomings. Pubmed and Google Scholar fall short, because searching by name just doesn’t work. Scopus and ResearcherID are nice and provide additional features, e.g. the Hirsch number or citations of your papers. But the commercial interests of Elsevier and Thomson Scientific have introduced important limitations.

So for the time being, the best tool is Nature Network. And there is great potential for improvements, e.g. integration with Connotea or linking to coauthors that are also Nature Network members.

In related news, Ian Mulvany today announced on his Nature Network blog that Connotea is now OpenID enabled.

The difficulties for German authors are much smaller, and this is probably true for other European languages. We do have a fair number of surnames, but umlauts in our names (e.g. in the very popular last name Müller) are frequently lost. As Cesar Sanchez pointed out in the Nature Nanotechnology discussion mentioned above, Pubmed started using diacritics, including umlauts, last year (Diacritics in PubMed® Displays and Searching). And the American Physical Society started to allow Chinese, Japanese and Korean authors to use names in their own language (Editorial: Which Wei Wang?).

But the problem is the same. There are at least 6 different M. Fenner in the MEDLINE database, one of them my cousin. Using the middle initial can help. I try to publish as MH Fenner, and very smart people will figure out that papers written by H Fenner are from my father. And what happens when you marry? My wife and I have different last names (which is uncommon in Germany) and one small reason was the scientific track record (including publications) connected to this name. Some of the other issues are nicely summarized by in a PLOS post by Richard Cave (Unique Author Identification, thanks Cesar Sanchez).

The solution? We need unique identifiers for authors. I recently wrote about ResearcherID, one such effort announced this January by Thomson Scientific. The problem with ResearcherID is that author identity is a very sensitive issue and many people will be reluctant to rely on a private company for that. Elsevier is doing something similar in their Scopus database (The Scopus Author Identifier).

Author identifiers should really come from a neutral organization such as CrossRef (a publishing organization that brought us the DOI to uniquely identify a scientific paper). They held a meeting with various interested parties in February 2007 (CrossRef Author ID meeting). Because of the number of parties involved and the different issues, this is a very slow process. And what is the U.S. Library of Medicine doing on this issue? After all, they publish MEDLINE, the most important database of biomedical research.

During the day I treat cancer patients and do cancer research at a German University hospital. But I have been interested in programming for many years, first with the Applescript language (I was involved with the Macscripter website) and more recently Ruby on Rails. Open sharing of resources and information using blogs and other tools is common in these communities.

I joined Nature Network last August. In May I had started a science blog on my on server, more as a personal experiment than a serious blogging effort. The inspiration came from an article by Laura Bonetta published in Cell entitled Scientists enter the blogosphere. Moving to Nature Network gave me a much wider audience, but also meant a commitment to a more serious blogging effort.

The topics I pick for my blog mostly relate to how the internet has changed the way we publish our scientific work. Writing regular blog posts about these topics has forced me to do a lot of background research and I have probably learned more than I would have by just reading papers and other people’s blog posts or comments.

In the early 1990s, when Internet use was mostly by dialup and could cost $10 per hour, I was involved with the Compuserve network. I was an assistant Sysop with the MacUser Forum and we had message boards, chat and file downloads.

Nature Network has many of the same features, but with the technology of the 21st century. The community of fellow NatWorkers is one big reason to stick around here. Not only did I learn a lot from others, but is also just fun to learn about singing scientists and houses in Cromer.

Did Nature Network change my life? At least a little bit already. I wish Nature Network a happy birthday.

Myelodysplastic syndromes are malignant disorders of the bone marrow that show insufficient production of blood cells and harbor the risk of progressing to acute myeloid leukemia. The 5q syndrome is a distinct subtype that is characterized by loss of genetic material at chromosome 5q31, a characteristic morphology of the bone marrow, and a fairly benign clinical course. We would expect to find a tumor suppressor gene in the deleted region on chromosome 5, but no genetic changes on the other allele of chromosome 5 could be detected. The traditional strategies to identify the genetic changes responsible for the 5q syndrome therefore didn’t work.

Research by Benjamin Ebert et al. presented at the American Society of Hematology meeting in December and now published in Nature³ claims to have finally identified the genetic change responsible for 5q syndrome. They used RNA interference to in turn knock out each of the 40 genes in the common deleted region on chromosome 5q31. They observed the phenotype of human hematopoetic progenitor cells transfected with short hairpin RNAs. The shRNA targeting the gene RPS14 recapitulated the phenotype of patients with 5q syndrome.

To confirm the role of RPS14 in 5q syndrome, RPS14 was overexpressed in hematopoetic cells from patients with 5q syndrome and indeed reverted the erythroid differentiation effect. Reduced expression or inactivating mutations of the one remaining RPS14 gene in 5q syndrome patients was ruled out by sequencing and gene expression profiling.

Further evidence for the importance of RPS14 in 5q syndrome comes from studies of related genes. Expression of RPS14 is required for proper function of the 40s ribosomal subunit. Germline mutations of two other ribosomal proteins, RPS19 and RPS24, have been identified in the congenital disorder Diamond-Blackfan anemia, and the disease in these children has similar features.

As a medical student, I was fortunate enough to listen to Alfred Knudson give a talk about is two-hit hypothesis of tumor suppressor genes², and I was fascinated by cancer genetics ever since. Genetics continues to be a driving force in our understanding of cancer. The next big step in the 5q syndrome story will be a better understanding of why the drug lenalidomide works so well in this disease, and how this relates to RPS14 function. More information about the Ebert paper can be found in this Nature News and Views article.

¹ Croce C. Oncogenes and Cancer

² Knudson AG. Mutation and Cancer: Statistical Study of Retinoblastoma

³ Ebert BL et al. Identification of RPS14 as a 5qsyndrome gene by
RNA interference screen

The recently launched Research Blogging is a related effort to aggregate blog posts about science.

Like many other science bloggers, I do not write directly about science, but rather about issues that are important for scientists. I have so far been reluctant to blog about science. Why is that? Jon Udell two weeks ago wrote a blog post where he observes that bloggers talk to bloggers and scientists talk to scientists. I agree with him that we need better tools to integrate scientific publishing and the blogosphere.

But more importantly, many Scientists (including myself) are still reluctant to talk about their science in the blogosphere. This is understandable if you have unpublished results, half-baked ideas, etc that are not yet ready to be shared with the scientific community. But non-scientists usually don’t read the scientific literature, as most papers are difficult to impossible to understand – and access to scientific literature can be expensive. We shouldn’t forget to communicate our research to the community. If we don’t do that, we shouldn’t be surprised if the traditional media either don’t report our research at all, or give a distorted view that we are not happy with.

Having said that, I don’t have have any immediate plans to blog about science. My area of expertise is cancer research, and talking about cancer research is even more complicated because of often unrealistic expectations that these findings quickly translate into new treatment options for patients.

The article has already been widely cited, including of course Nature News and Nature Network, but also Noble Intent and DigitalKoans.

The original paper by Errami and Garner was published in Bioinformatics. They used the search engine eTBLAST to find duplicate papers and deposited the results in a database called Deja Vu. When you search Deja Vu for scientists you know, you find scary results.

Nobody likes duplicate papers, but it’s just another trick to improve your publication record. Another popular trick is the inflation of paper authors. But the basic trick is still the heavy use of the least publishable unit or LPU.

Both authors and journals are inclined to publish as many papers as possible. So what will change these practices? If the quality of the work of a scientist isn’t simply measured by numerical indices such as number of publications. So it’s up to those that decide about grants and jobs to find a better way to pick up the best scientists.

Corie Lok has started a discussion about this topic in the Publishing in the New Millenium Forum, please post your comments there.

Not many institutions in Germany have the resources (or vision) to hire such a person. I was therefore glad to see a job posting for a scientific writer at the Max Planck Institute for Infection Biology in Berlin, Germany. More institutions, including of course my own university, should follow this example. (Disclaimer: I’m not affiliated with the Max Planck Institute in any way. I worked across the street for five years).

The missing unique ID for authors makes it impossible to automate the creation of publication lists for authors or insitutions. You already can get email alerts or RSS feeds of papers published by a specific person (colleague or competitor), but without unique ID, this is tricky. ResearcherID will also automate the maintenance of your Hirsch number (for those interested in this kind of metric).

Scientists interested in ResearcherID should register at the reseacherID.com website. Registration is currently by invitation only.

There is one major problem with ResearcherID. It is not clear from the press release of who owns and controls this information. This is a very sensitive issue and Thomson Scientific a major player in this field. Remember Microsoft Passport? The majority of web users was uncomfortable having a central user ID and password managed by Microsoft. The service eventually failed, and we now have the open standard Open ID. Similarly, ResearcherID can only work if the user database is either shared openly or hosted by someone like the NIH.

But Microsoft Office 2008 for Mac is probably the most important new product from a scientists perspective. If you live in the U.S., you can buy Office right now, us international users will have to wait a few weeks. Office 2008 runs natively (i.e. much faster) on Intel Macs, so everybody with those newer Macs wants to upgrade for that reason alone. Powerpoint and Excel are important, but Microsoft Word is the application I use most of the time.

Microsoft Word 2008 has a new feature called citations, basically a simple reference manager. This feature first appeared in Office 2007 for Windows. Citations integrates nicely with your Word document, but can’t handle the more sophisticated needs of a research paper, e.g. automatic importing form online databases and dozens of reference styles for all the journals you possibly want to submit to.

Unfortunately you can’t use the current version of Endnote with Microsoft Word 2008. Thomson Scientific (the Endnote publisher) is working on a version for Microsoft Word 2008. The update is made more difficult by the fact that support for Visual Basic for Applications (VBA) was dropped in Office 2008. The alternative reference managers Bookends and Sente also currently don’t work with Microsoft Word 2008.

Microsoft Word 2008 by default uses the new XML-based .docx format introduced in Office 2007 for Windows. This format was originally not supported by Nature, Science and other STM publishers (read this Nascent post for the reasons behind it). Nature is now able to accept Word 2007 (and Word 2008) files, but this might not necessarily be true for other publishers.

The publisher and editors of the journal decided to make this step to have the journal articles better indexed in databases such as PubMed and available to more readers. This should lead to more citations of journal articles, resulting in a better Impact Factor and reputation of the journal.

100 years ago, German was an important scientific language, but now only 2% of journals indexed in Medline are in German. I personally haven’t written a scientific paper in German in for many years. I sometimes regret that I can’t use German (or French or Italian for that matter) to report by findings or express my ideas. But in the end it makes the exchange of ideas between scientists much easier if we can all use the same language. And Nature Network is a good example for this.

The draft article was posted online last week. In true Web 2.0 spirit, M. Mitchell Waldrop has invited readers to post comments and promised that he will use them for the final print version. I wrote a comment and suggested that the increasing role of Web 2.0 companies such as Google, Microsoft and Adobe in the online creation and distribution of science should be mentioned in the article.

I haven’t seen this information reproduced somewhere else, but the number for most of the other Google products were higher than 2006, as expected. And I don’t have the numbers of searches in Google Scholar compared to PubMed (you would have to buy this information from companies like comScore). But does this indicate that there is something wrong with Google Scholar? As a PubMed user for 15 years, I still like to get search results chronologically and not by (perceived) relevance. And I like the detailed search options and integration with other databases.

Other search engines in this field include Scirus from Elsevier and Windows Live Academic Search from Microsoft. Incidentally, TechCrunch (and others) reports this week that Microsoft is about to acquire Fast Search & Transfer, the Norvegian company that provides the search technology to Scirus.

Some big company names, but for me Pubmed is still the first stop when searching for journal articles. And PubMed will become more important as the NIH now requires the deposition of all NIH-funded research papers in PubMed Central.

Impact factors have been around for more than 40 years and they generally been very helpful. But there are two big problems:

Impact Factors are published by one privately owned company
Given the importance of Impact Factors for many aspects of scientific publishing, it would be preferable if there were alternatives. And Impact Factors are not freely available, but must be purchased from Thomson Scientific.

Impact Factors might not be the best tool to measure scientific quality
Impact factors have several shortcomings. Because they are a convenient way to judge the scientific output of a person, organization, journal or country, they are often overused. They should for example not be used to compare journals in different fields, e.g. cell biology and particle physics. Measures like the Hirsch Number might be a better tool to measure the scientific output of an individual scientist. And sometimes the judgement of your peers in the field is more important than simple numbers.

The SCImago Journal Rank indicator tries to overcome these two shortcomings. The index was created by the SCImage Research Group, located at several Spanish universities. The index uses information from the Scopus abstract and citation database of research literature owned by Elsevier.

In contrast to the Impact Factor, the SJR indicator measures not simply the number of citations per paper. Citations from a journal with a higher SJR indicator lead to a higher SJR indicator for the cited journal (more details here). This approach is similar to PageRank (described in this paper), the algorithm for web searches by Sergey Brin and Lawrence Page that made Google what it is today. Eigenfactor is another scientific ranking tool that uses a PageRank algorithm.

Most of the time, journals with high Impact Factors have high SJR indicators. Nature and Science are still head to head. We will find unexpected results and discrepancies between the two over time. In my field of oncology, both the Journal of the NCI and Cancer Research are ranked higher than the Journal of Clinical Oncology.

You can read more about the SJR indicator in this Nature News article.

The Open Access mandate for NIH-funded research was voluntary since 2005. Fewer than 5% of research papers were actually made publicly available. The process and discusson about making this requirement mandatory as part of the Fiscal Year 2008 Labor, HHS, and Education Appropriations Bill was going on for many months (as reported previously by me and others), including a lobbying effort to stop this mandate called PRISM.

A large part of biomedical research is funded by the NIH and this change in NIH policy will probably have a big impact on how most biomedical journals do their business. A wonderful christmas present for all scientists.

Good
Most importantly, there is only one version of your manuscript. You and your coauthors can simultaneously work on different parts of the manuscript, the software saves all versions of the text. Buzzword is very pleasant to use, thanks to the elegant interface. Importing and exporting to Microsoft Word is possible. And Buzzword is free.

Bad
Buzzword doesn’t handle references. You can create footnotes, but for referencing papers, you have to use another software. That’s what I did in the final stages of manuscript writing, switching to Microsoft Word and Endnote. Buzzword doesn’t show you the changes between different versions of a manuscript. This makes it difficult to see the changes that your coauthors made. Because Buzzword is an online word processor, you can’t use it without an internet connection. One of my coauthors had Buzzword crashing several times, probably because of a slow internet connection.

Conclusions
Manuscript writing with Buzzword is fun. But without support for references and better version management, it is not ready for prime time.

But when the just published paper comes to the opposite conclusions, and has done all the right experiments, your work can become almost worthless. This just happened to me with a paper in the December issue of Nature Medicine about the role PPARgamma in osteoclastogenesis. The group of Ronald Evans used mice with a conditional knockout of PPARgamma in osteoclasts to study the role of this nuclear hormone receptor in bone diseases. Their elegant work clearly shows that PPARgamma promotes osteoclast differentiation and activation.

Unfortunately we have very different results in a tissue culture model of osteoclast differentiation, consistent with previous publications. But it is hard to argue with a mouse model. We will carefully look at the data to decide where we go from here. But for now it’s just frustrating.

The workshop was organized as a 1 1/2 day course with many group exercises. We learned about good scientifc language, tables and graphs and other typical aspects of manuscript writing. Stuff you will also find in books such as this one. But a workshop is certainly a better learning experience – and more fun.

To my surprise the most important lesson was something else. Good scientific writing always has a message, targeted towards the audience. This is probably obvious for anybody with basic experience in journalism. But it is easier said than done when the results of your experiments are complicated and not easily fit into a model. I will try much harder to keep this in mind, e.g. by making sure that the abstract and tables/figures can speak for themselves.

I would like to make another suggestion. The quality of a research paper should not only be judged by the number of citations it receives (which improves the Hirsch number of the author), but also by the number of authors. There are of course research projects that are only possible with large numbers of collaborators, but many biomedical papers probably only need 2-4 authors, but rather have 5-8 authors.

One good rule of thumb is a number of papers published per year. If that number is too high (e.g. more than 10), than the author has probably not contributed significantly to all those papers. Department heads often fall into this category. These rules can easily be applied when reviewing job or grant applications.

One of the most famous examples for authorship not perused is the 1922 CMAJ paper about the action of pancreatic extracts on blood sugar in diabetics. John MacLeod was not a coauthor, but rather thanked in the acknowledgements. As head of the department he provided mainly logistical support. He still went on to win the Nobel Prize for the discovery of insulin the following year.

SciGen is of course a hoax. There are other famous hoaxes in science, including the 1996 Alan Sokal paper “Transgressing the Boundaries: Towards a Transformative Hermeneutics of Quantum Gravity”.

What do these hoaxes have in common? They randomly generate pseudo-scientific language. Important ingredients are buzzword frequently used in the field and standard phrases. If we look carefully, we find examples of this random-talk in our own work as well.

I simply like the new name. The word Goobledygook was invented by the U.S. congressman Maury Maverick and used in a New York Times article published May 21, 1944. Maury Maverick was fed up with the vague, pompous and repetitive language used in Washington at the time and invented this new word for it. He was fighting against the all too common use of Gobbledygook. The following text is from his New York Times article:

Plain and simple speech appeals to everyone because it indicates clear thought and honest motives. Here is the point: anyone who is thinking clearly and honestly can express his thoughts in words which are understandable, and in very few of them. Let’s write for the reader and not for ourselves. Make the writing do what it is intended to.

This text might as well be from a recent editorial about scientific writing. We had a very lively discussion about this topic in the Ask the Nature Editor Forum. I first came across the word Gobbledygook when I wrote about readability tools. The Nature Literacy Trust has created such a tool and called it SMOG or Simplified Measure of Gobbledygook.

Papers is a wonderful piece of software that helps to organize the clutter of PDF files on your hard drive. My wish for version 2.0: integration of RSS feeds from journal table of contents or custom PubMed searches.

For more information, visit the Papers Group on Nature Network.

In biomedical research we need (mostly greek) symbols, but for mathematicians and physicists the needs are far more complicated. The final version of the STIX fonts should be released soon, and the fonts will be freely available. I wrote previously about the lack of symbol font support in online writing tools such as Google Docs and Buzzword. The STIX fonts should be a tremendous help for them.

To improve this situation, the European School of Oncology – where Kathy Redmond is coordinator of the media program – last year started an annual Best Cancer Reporter Award. And two days ago, ESO organized a media forum entitled Cancer: Time for a Reality Check in Rome. And ESO is partnering with Nature Clinical Practice Oncology to collect summaries of important research findings to journalists.

For those of us working in cancer research, it is important to remember to communicate our research findings not only in journal articles and scientific meetings. We probably have to do a much better job in talking to the media and the public. One example would be to start a blog about a particular area of cancer research or cancer patient care. The number of quality blogs in this area could be much higher, and some blogs even had to close down.

The arguments used by the German science minister Bernhard Rust at the time are not worth repeating – they are ideological rather than scientific arguments. But the story is interesting for today’s scientists because it serves as a reminder that free access to scientific knowledge means much more than the open access vs. closed access debate.

This is an important announcement, because the MPG is one of the largest research organizations in Germany and Springer the second largest STM (Science, Technology, Medicine) publisher after Elsevier. Just a few weeks before this announcement, both the Howard Hughes Medical Institute and the University of Göttingen announced that they will pay their authors for Open Access in Springer journals (Springer Open Choice).

Time will tell whether these announcements are coincidences or the part of a larger trend from a subscriber-pays to an author-pays model.

But what is good scientific writing? Two weeks ago I suggested a few good books on the subject in a blog post. Once you have written the paper using the advice in these books, you can use a number of tools to measure the readability of your paper.

Flesch Reading Ease Score:

206.835 – (1.015 x ASL) – (84.6 x ASW)

ASL is average sentence length and ASW is average number of syllables per word. The Flesch Reading Ease Score can be between 0 and 100, 100 being the most difficult.

The Flesch-Kincaid Grade Level Test is similar, but rates the text on a U.S. school grade level. Both Flesch scores are built into recent versions of Microsoft Word or you could use the Open Source application Flesh. Flesh is much easier to use than the Microsoft Word tool and will also open PDF files. Try to rewrite your manuscript if your Flesch Reading Ease Score is too low, e.g. below 30.

Readability is important not only for manuscripts. Informed consent forms for patients wishing to participate in clinical trials in Medicine are often difficult to read. And a perspective article by Celeste Condit in the current Nature Reviews Genetics talks about How geneticists can help reporters to get their story right. She points out the importance of readability to communicate often highly technical material to lay people.

Microsoft Office Live Workspace allows the sharing of Microsoft Office documents. Just as with Adobe Share, and in contrast to Google Docs, you can’t directly edit documents online.

Why should a scientist care about all this? Because these are excellent tools to share your presentations and to collaborate on manuscripts you are writing. No more emailing Powerpoint, Excel, Word or PDF files back and forth. And with several major players (Microsoft, Google, Adobe and the smaller Zoho) offering web-based document sharing, this will soon become as commonplace as web-based email.

In related news, Adobe today also announced the acquisition of Buzzword, the online word processor that I wrote about last week.

A few years ago I bought an old copy of Scientific Writing by Lester S. King and Charles G. Roland (Journal of the American Medical Association 1968). King and Roland encourage the reader to avoid unnecessary verbosity and use the active voice more often (they don’t talk much about graphs and statistics). Scientific writing seemed to have almost the same problems 40 years ago. This raises two interesting questions:

• Could the writing of the average scientific paper be improved and are we in need of these books?
• Does the quality of the writing help in getting a paper accepted? In other words, should I make the effort and improve my writing skills or should I rather do experiments?

Journal editors can probably better answer these questions. My guess would be that most people would answer these questions with yes and no. So we will continue to struggle with our writing skills, especially those of us who learned English as a second language, but it will not be that important in our career. I therefore rather list some other books about scientific writing that I own and can recommend:

• Medical Writing: a Prescription for Clarity by Neville W. Goodman and Martin B. Edwards (Cambridge University Press 2006). Very detailed instructions about the use of the language. Includes many examples of bad scientific writing and suggestions for better alternatives.
• The Elements of Style by William Strunk Jr. and E.B. White (Logman 2000). The classic text from the beginning of the 20th century, not specific to scientific writing.
• The Visual Display of Quantitative Information by Edward R. Tufte (Graphics Press 2001). Another classic text and a beautiful book.
• Clinical Epidemiology by R. Brian Haynes, David L. Sackett, Gordon H. Guyatt and peter Tugwell (Lippincott 2006). This epidemiology/statistics textbook has a small chapter about scientific writing. More importantly, it is very helpful in how to read a scientific paper and the statistics behind it.

And I finish with an example for good scientific writing that is used a lot in the books mentioned above:

We wish to suggest a structure for the salt of deoxyribose nucleic acid (D.N.A.). This structure has novel features which are of considerable biological interest…

This is of course the opening paragraph of the 1953 Nature paper by Watson and Crick. Which brings us back to Cambridge…

The topic was of course very different from say a meeting about RNA interference, but many things were surprisingly similar. But I found a few differences very interesting, and I believe a scientific meeting could learn from them.

Many of the talks were about ongoing work that hasn’t been finished yet. Ruby on Rails is Open Source software, so there is less worry about intellectual property and people talked freely about their ideas and problems. This is in contrast to the typical meeting in my field (oncology), where only the most established researchers are not afraid to present data that have not yet been published or at least submitted. This of course brings us back to the Open Science discussion both on Nature Network and elsewhere.

Partly because of this openess, but also because it was a technology meeting, the conference is much more accessible for those that couldn’t make it to Berlin. Not only are the slides of most presentations available for download, but there are also numerous blog posts. We’ll soon also see podcasts and videocasts (the keynote lecture is already available). I’m a big podcast fan and I would love to see more of this from the conferences in my field.

In contrast to articles in Elsevier journals (which become freely available after 6 months and Elsevier holds the copyright), articles in Springer journals will be available immediately and the author will retain the copyright.

I’m glad to see this agreement happen. It probably makes sense for Springer, HHMI and, most importantly, the involved scientists. Even more so now with the charged discussion about PRISM.