<br><br><div class="gmail_quote">On Mon, Feb 25, 2013 at 12:24 AM, Andrew A. Adams <span dir="ltr"><<a href="mailto:aaa@meiji.ac.jp" target="_blank">aaa@meiji.ac.jp</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="im">>> On your point on central deposit, I beg to differ, as you<br>
>> know. Deposit locally then harvest centrally is far more sensible<br>
>> than trying to mandate different deposit loci for the various authors<br>
>> in an institution.<br>
<br>
</div>Peter Murray-Rust replied:<br>
<div class="im">> This is not axiomatic. The protein community requires authors to<br>
> deposit sequences communally - and they do. The genome community<br>
> requires genes deposisted and they do. The crystallographers require<br>
> crsytal structures and it's 100% compliance. The astronomers...<br>
<br>
</div>"The community requires"? How, exactly?<br></blockquote><div><br>The International Union of Crystallography requires that complete data (as complete as possible) for experiments are made available at time of submission and subsequently made public at time of publication. They have argued this consistently and all reputable journals now mandate this. This is because journal editors recognize the importance. If it's a protein it goes in the global Protein Data Bank.<br>
<br>The same happens throughout molecular reductionist bioscience. The data in the human genome has created nearly 1 trillion dollars of public value (Battelle's figures).<br></div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
I do not dispute that there are a smal number of subfields where OA of papers<br>
has been successful without mandates, and in some areas instead of in<br>
addition there is deposit of certain types of research data unmandated.<br>
However, they are a tiny minority of academia. Do you disagree with this<br>
assessment?<br></blockquote><div><br>If you are saying that most science is badly and sloppily published I'd agree. If you are saying that a significant fraction are actively working to improve this all the time by requiring publication of the complete experiement. I'd agree. But you are labelling these as "a tiny minority" (with no justification). This is simply not true - there are a very large number of scientists who make their data available. But even if it were a "tiny minority" (and that was true of crystallography 40 years ago) that does not excuse the sloppiness of the "vast majority". <br>
</div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
The question then becomes how we get the rest of academia to do so. Despite<br>
the possibilities having existed for over twety years, the vast majority have<br>
yet to do so, despite it being in their interests.<br></blockquote><div><br>Well, last week we launched a program to build a Semantic Web for Materials Science - an area where > 1 billion dollars of data could be easily captured if we had the will. If people like you praise and support these initiatives it helps. <br>
</div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
Who can require them to do so? Their employers and funders.<br></blockquote><div><br>There is a concept of scientific method and ethical behaviour which most scientists adhere to strongly. That is ultimately the strongest driver. Good scientific practice becomes part of the funding policy not for religious reasons but because it is inherent in good science. That's why research councils are mandating data management. Because it's essential for goo science. <br>
</div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
What is the most efficient way for employers and funder to mandate deposit (a<br>
mandate requires at least some level of compliance checking otherwise it's<br>
really just a suggestion).<br></blockquote><div><br>The most powerful comes when the scientific societies create protocols to support good science. For example the Am Soc Bio Mol Bio (ASBMB) 5 years ago said scientists must publish their proteomics mass spectra. (BTW without these data the "final PDF" is unverifiable and often meaningless". The naysayers said that authors wouldn't comply and would go to other journals. The ASBMB insisted on data for Mol Cell Proteomics. Within 2 years all the other journals followed suit and it's no mandatory.<br>
<br>That's how good science works.<br><br></div><br clear="all"></div><br>-- <br>Peter Murray-Rust<br>Reader in Molecular Informatics<br>Unilever Centre, Dep. Of Chemistry<br>University of Cambridge<br>CB2 1EW, UK<br>
+44-1223-763069