Mitsuteru N › Likes

What kind of platforms are you thinking about here? - Egon Willighagen

CKAN, FigShare, Dryad (via EvoMRI) and the like. It's for a social computing community. - ReaderMeter

Anyone aware of a a feature comparison of data repositories? #opendata - Daniel Mietchen

"Dryad accepts data in any format as long as it is associated with a primary publication" I guess this is too restrictive as it makes the platform unsuitable for hosting research datasets the Wikimedia Foundation generate but that do not make the object (at least not immediately) of scholarly publications. - ReaderMeter

Not an actual feature comparison but at least a useful directory to start from: http://wiki.civiccommons.org/Data_Pl... - ReaderMeter

I started a basic feature comparison spreadsheet at: http://bit.ly/OpenDat... - ReaderMeter

Well papers are academic currency. Any way to increase your wealth (utility be damned) :) - Rajarshi Guha

But three? Within six months? 200% inflation rate is going to kill this system... :) - Pawel Szczesny

@Pawel... agreed. This is the same paper. It is not uncommon for biology groups to publish the 'tool' separetely from the 'science', but this sounds ridiculous... plagiarism it is... actually, all journals I have been reviewing for in chemistry, do not allow results to be published before... I can't believe there are so many angles to this tool that those journals would have allowed it... within 6 months... that means they must have been submitted simultaneously :) - Egon Willighagen

Actually they're doing themselves a major disservice. By publishing the same thing 3 times they effectively divide their citations by 3, which harms their H-index. - Paul Gardner

And none of these 3 papers cite Jmol! Or even mention it... - Egon Willighagen

NAR often includes previously published databases and software. - Matt Hodgkinson

I've also heard of a rejection to the NAR webserver issue b/c of a Bioinformatics Application Note. But I'm not sure if this is a general policy. Once you're in the NAR db / webserver issue, you can re-submit after 2 years. - Michael Kuhn

I'll play devil's advocate. Apart from the reaction against CV stuffing is there any good reason not to do multiple publications for a service? If the argument were, for instance, to reach a series of different audiences? - Cameron Neylon

multiple pubs in multiple venues are fine. But pubs are currently a currency and basis of competition (amongst other things); from this POV, spamming journals with multiple articles devalues the individual articles - Rajarshi Guha

Agreed but surely its the author's choice to balance that devaluation against potential value gain of reaching new people? I guess what I find interesting is that people feel that protecting against publication inflation is a bigger concern than getting information out efficiently. Similar case where a piece in PLoS Currents was subsequently published elsewhere and everyone got their... more... - Cameron Neylon

I checked the website, and the have an attribution clause... I could not find the attribution requirements, but nothing stops them from asking people to cite all *3* papers... - Egon Willighagen

@Cameron... I think it's a problem of inflation, and devaluation. 3 papers is simply more rewarding, and everyone not publishing more or less the same thing trice is effectively punished. - Egon Willighagen

Perhaps, but is that not a symptom of measuring the wrong thing? If we actually measured re-use (e.g. citations) and three papers meant the number of citations were cut in three for each paper and the total number was the same then we'd be ok right? No devaluation? The problem here is not that its being published three times but that we value the wrong things (number of papers) in a system that enables (or even encourages) cheating. - Cameron Neylon

What @egon said. My basis for this argument is that, in principle, multiple pubs in different venues are fine (I'm not sure how different the venues were for this case). And in a world where the nuances (or lack thereof) of these multiple pubs are taken into account, this would be fine. But in the real world, where jobs/grants/promotions are (unfortunately, frustratingly) based on a... more... - Rajarshi Guha

@cameron - absolutely! We are measuring the wrong thing. But, that's what we're measuring. So to stay in the race, we (well, not me, it doesn't matter to me much anymore) play the game, whose apparently best strategy is to publish as much as we can. I'm sure that with your and others' efforts this will change one day - but people still want to get their jobs/grants/promotions ... - Rajarshi Guha

Agreed - and this isn't a case where I'd argue much in their favour. But the thing with PLoS Currents was a bit different but got a very similar response. Interested whether people feel that's as egregious a case. - Cameron Neylon

is there a link to the PLoS Currents discussion? - Rajarshi Guha

Not sure if these particular 3 papers are what I usually think of as duplicate papers. The "Acta Crystallographica Section F" one is part of a special issue about the JCSG pipeline, so I think it's reasonable there even if it's duplicating things. And my opinion is that the NAR database/server issues are also a special case - as they provide a resource to the community and often describe websites that have been published elsewhere. In short, not the most straight forward example of duplicate publications. - Mickey Kosloff

Cameron, if you're playing devil's advocate, don't forget to send an invoice to NPG, because they will profit the most from perceived inflation of papers outside of Nature* ecosystem. :) But let me play the game as well - if we allow for such marketing strategy, it gives yet another advantage to people who use English natively and have no problems to write five different stories on the same discovery. Yet another penalty for not being British? Thank you so much, Cameron ;). - Pawel Szczesny

Pawel++ - Egon Willighagen

Mickey, while I agree these are "special cases", not clear duplications, I still don't really get it why it's allowed in a first place. When I was reviewing a manuscript for NAR special issue I'd asked authors to improve the service in comparison to the original (published few months earlier) despite clear policy on allowing duplicates. Today, probably I would refuse to review for NAR special issue at all... - Pawel Szczesny

:) - Rajarshi Guha

Good point about the need for better search and analysis tools - Mr. Gunn from YouFeed

He's braver than me....not sure I want to go there. I did share my Paternal Haplogroup and projections of what I might look like fat, bald and old though: http://tinyurl.com/5upk6p9 - Antony Williams

@Egon, I'm going to wait for the publication of this article before talking about it :-P - Pierre Lindenbaum

Congrats!! - Björn Brembs

I'm glad you got a paper out, but I'm sorry you couldn't get it published in a real journal. - Bill Hooker

@Pierre... ah, embargoed... obviously increases dissemination of scientific results :) - Egon Willighagen

No, not yet. But you can use the SPARQL end point to 'download' all the data... - Egon Willighagen

I was trying to see how to format the solubility data. - Andrew Lang

That's next up... Bioclipse has a manager that creates RDF... see: http://chem-bla-ics.blogspot.com/search... - Egon Willighagen

I will use CHEMINF, which is a bit cryptic (inherited from OBO), and elaborate (deliberate, to allow proper provenance of data)... paper should be submitted any day now... - Egon Willighagen

Examples: http://code.google.com/p... - Egon Willighagen

CHEMINF and related ontologies also allow marking something up as solvent... for that, browser the HTML+RDFa linked to here: http://chem-bla-ics.blogspot.com/2010... - Egon Willighagen

(I think I added it there... need to check, actually...) - Egon Willighagen

Andrew, the solvent annotation is part of this page: https://hudson.ch.cam.ac.uk/job... - Egon Willighagen

Thanks Egon. - Andrew Lang

Does this work? <?xml version="1.0"?> <rdf:RDF xmlns:rdf="http://www.w3.org/1999..." xmlns:chem="http://www.blueobelisk.org/chemist..."> <rdf:Description rdf:about="http://old.oru.edu/cccda..."> <chem:Solute>benzoic acid</chem:Solute> <chem:Solvent>1,4-dioxane</chem:Solvent> <chem:concentration>3.088</chem:concentration> </rdf:Description> </rdf:RDF> - Andrew Lang

Unfortunately the RDF file that Pierre created for our solubility data is now a broken link http://usefulchem.blogspot.com/2008... - Jean-Claude Bradley

@Andrew: sure. There are no rules, in that respect. The use of common ontologies is merely a matter of convenience... - Egon Willighagen

How's this? Is this useful? Would you add/change anything? http://showme.physics.drexel.edu/onsc... - Andrew Lang

That's very XML-like... have a look at this: https://gist.github.com/766302 - Egon Willighagen

Bioclipse can generate RDF directly from the ONS Google Spreadsheet, but I have not had time yet to look at that... will try that soon. - Egon Willighagen

Thanks Egon. I don't see a closing to the <chem:Measurement> or an opening for </rdf:Description>. Should the </rdf:Description> be </chem:Measurement>. Also the units are M. Would it be ok just to stick that at the end of the number, e.g. 2.380M, or do I need to put units separately from the value? - Andrew Lang

Yes, the rdf:Description should be a closing chem:Measurement. You can better change the predicate to molarConcentration, I think. That way, the field stays a float. - Egon Willighagen

Thanks Egon - Feed updated: http://showme.physics.drexel.edu/onsc... - Andrew Lang

I'm trying to convert the SMILES to InChIs. I've tried http://cactus.nci.nih.gov/ and ChemSpider and both fail and are slow. Also some urls generated don't resolve, e.g.: http://rdf.openmolecules.net/InChI=1...) - Andrew Lang

Andrew, there should be a ? just before the InChI= ... like: rdf.openmolecules.net/?InChI=1/CH4/h1H4 ... - Egon Willighagen

This could be really useful for us too - we need the SMILES to be in a particular order for the starting materials to generate libraries of Ugi products http://usefulchem.blogspot.com/2010... - Jean-Claude Bradley

Exactly. I'll be looking into Andrew's code after Christmas, and help you out if need be. - Noel O'Boyle

@Neil - Cool, igraph is new for me, I learned something! - joergkurtwegner

Got to cechkout Gephi - Cytoscape is really slow for graph vis - Rajarshi Guha

wow .. where did that come from, sexy networks :) thanks - Pedro Beltrao