Depth-First - Latest Comments

Re: Free Access to ACS Publications and Why You Can't Have It (Yet) | Depth-First

Timo — Tue, 21 May 2013 03:28:24 -0000

May 21st - I can not access any of the publications from the links above whithout logging in to ACS publications. As I am not willing to register there, I can not see whether the papers would be available for me or not - as it is, they are not...

Re: The Horrifying Future of Scientific Communication | Depth-First

James Ashenhurst — Fri, 10 May 2013 22:08:06 -0000

While it might be, I'm not sure that this is a classic innovator's dilemma, in the example of email/snail, netflix/blockbuster, napster/music labels, craigslist/classifieds, or even hydraulic excavators vs. steam shovels. Many of the above had advantages in providing the same information albeit in a cheaper and faster manner.

My opinion on two questions, and then the central problem.

1. Would radically cheaper journals have a significant effect on established publishing?

While journals *are* expensive, they are not experiencing (to my knowledge although I would be happy to be corrected) significant problems with piracy that might lead to downward price pressure, although this would be extremely easy to do. Secondly, since costs are borne indirectly (i.e. as one of many contributions to "overhead" in the primary users' research grants, but not as a discrete item) they are more abstract.

2. Would journals that allowed for significantly *faster* publishing have a significant effect on established publishing?

Not sure that it would, at least in the field of chemistry. All but the most exciting results in the chemical literature are of the "interesting! I'll file that away for future use" category, rather than being of immediate utility. While there is incentive for "planting the flag" in the case of truly groundbreaking results, it does not (to me) seem overly significant to be able to publish results 1-2 months earlier than currently - especially with the advent of ASAPs.

3. The status problem for wannabe disruptive journals

As you pointed out, the essential problem for online journals such as PLoS one are those of status, and this is the key driving force in scientific publishing. Existing journals have extremely strong brands. Tell a complete scientific stranger you published an article in Science, and they will, at the very least, emit a respectful grunt. This is, of course, because journal space is scarce, rejection rates are high, and passage through the gauntlet means that the paper has passed through the (hopefully) skeptical eyes of top people in the field before publication who have every excuse to reject shoddy work. All this is a prelude to the fact that the high status of certain journals is not tangential to the scientific enterprise; it is central to the decision making process of tenure committees and to the awarding of grant money. Like it or not, this status is the medulla oblongata of science. Faster and cheaper is secondary.

How could "disruptive" online journals make an impact? I could forsee two ways, one practical and one fanciful.

Practically, there is room out on the margins. Libraries pay a ridiculous amount of money for low signal to noise journals like Tet Lett and Chirality. There is an opportunity for an intelligent, highly focused effort to pick off these stragglers. One possibility is to chop down the least publishable unit from a "communication" down to a reaction or two - allowing for the publication of an interesting dead end that might otherwise just end up in someone's Ph.D. thesis [or a full paper, which are on the decline these days]. This would make for a small time commitment from high-status research groups at very little risk.

A second, more fanciful idea would be that a high profile band of researchers broke off and published their own boutique scientific journal according to their own standards.

The bottom line is this - until high profile research groups at, say, Harvard, Princeton and Yale, among others, are seen to be publishing in these venues, they will suffer from a status problem. The challenge for journals like PLoS One is to obtain some degree of participation from high-status groups and leverage that into something bigger.

Re: The Horrifying Future of Scientific Communication | Depth-First

Rich Apodaca — Thu, 09 May 2013 11:24:54 -0000

the sticking point for me continues to be the communities that only count publications in journals.

A lot of this seems to lead back to the basic problem of "objectively" evaluating the outcome of an investment in research. Numbers of papers and the prestige of the journals they appear in are often used as a proxy given that few besides the scientist him/herself and a few peers can fully understand the significance of what was done.

So I'd have two responses:

(1) Disruptive technologies often exist side-by-side for some time with the mainstream technologies that end up vanishing (think minicomputers and personal computers);

(2) There are many ways to reward good research outcomes besides simply counting papers. For example, consider the enormous value companies now derive from ranking high on Google, compared to the widespread apathy toward search engines that preceded Google's disruption.

Re: The Horrifying Future of Scientific Communication | Depth-First

Rich Apodaca — Thu, 09 May 2013 11:03:34 -0000

Agreed. On the other hand, I recently had a discussion with a chemist who complained about the production requirements for Jove. If anyone else has had firsthand experience with producing a video for Jove, I'd be interested in hearing about it.

Re: The Horrifying Future of Scientific Communication | Depth-First

Rich Apodaca — Thu, 09 May 2013 10:56:07 -0000

A (well-written) scientific paper is far more than a data dump: it's a narrative, and reading papers outside your own area is a good way to build general knowledge. Reviewing raw data, experimental protocols, and quick conclusions are complementary to this, but serve other purposes and do not fully replace it.

Good point. Our current system (at least in chemistry) combines the experiment and interpretation into one package. The problem is: what happens when your experiment doesn't fit well into a narrative? Efforts to separate the two haven't gained a lot of traction, but there have been a few attempts. New services might make that feasible for the first time. The result could be... interesting.

Re: Chemical Structure Copy and Paste Problems | Depth-First

Igor — Wed, 08 May 2013 23:16:17 -0000

Rich,

I am not going to say that this problem is anywhere close to be solved but

"there is another way"...
http://www.chemaxon.com/marvin...
On Windows you can also use Accelrys Draw.
The copy/paste method used to work with a plugin script with ChemDraw but I have not tested it recently as I don't use that software anymore.

At least for some of better quality images this way might work for 90% of the time nowadays -
https://sourceforge.net/apps/m...

Re: The Horrifying Future of Scientific Communication | Depth-First

David — Wed, 08 May 2013 16:16:09 -0000

the sticking point for me continues to be the communities that only count publications in journals. The reason I countinue to publish is that (a) the NIH asks for list of relevant publications when I apply for a grant, (b) when a grant ends, success is viewed as a number of publications, (c) when you apply for a gordon conference, they ask how many publications you've had in the area in the last 3 years. For others, the raw count of publications may matter in hiring and tenure decisions.

Since you point to Hacker News, I'll point to Paul Graham's essays as my ideal method. They're well thought through, many have a list of 6 or 7 people at the end who have read drafts and presumably provided feedback and suggestions (e.g. http://paulgraham.com/startupi... ), which is really what I sometimes find valuable in peer review, and his essays are widely read by the tech start-up community.

If you're an established researcher, switching to this method may be accepted, but for a young person, it is hard to write on your NIH CV that you just put your results on your web site. And can you get the NIH to take blog posts from your site and put them on Pubmed Central or even have them indexed by PubMed? doubtful. And this results in limiting the established researchers, who often work with graduate students and post docs who need the long publication list to get a job.

I really hope we can break out of this, but I'm still clueless as to how.

Re: The Horrifying Future of Scientific Communication | Depth-First

Andrew — Wed, 08 May 2013 16:04:17 -0000

Very thoughtful piece. I think it's missing one point though.

A (well-written) scientific paper is far more than a data dump: it's a narrative, and reading papers outside your own area is a good way to build general knowledge. Reviewing raw data, experimental protocols, and quick conclusions are complementary to this, but serve other purposes and do not fully replace it.

This is not to say that the status quo is perfect or even particularly good for all the reasons you've highlighted, just that any replacement will have to include this (at least in some fields, including organic chemistry).

Very interested to see how the potentially disruptive platforms you've listed could affect things. It's not clear how, say, arXive would really change anything in organic chemistry - it's not going to promote reproduction.

What's definitely needed (and may or may not be disruptive) is what Michael Eisen referred to: a way of accountably annotating papers at the point of delivery, so that researchers don't have to waste time or rely on lab lore in order to find out a given paper is bunk. This would come with its own problems (who can comment? If it's limited, how does one gain this privilege? etc) but would be a step forward.

Re: How to Get a Job Like Mine | Depth-First

Brock — Wed, 08 May 2013 15:50:27 -0000

Sorry Rich -- I disagree.
I agree to be a good Scientific Software Engineer, you first need to have good Science (e.g. PhD in chemistry or biology, physics, etc) and understand the Domain (e.g. Drug Design).
But there is a LOT more that goes into making good software than just reading up on coding syntax. To really do good Scientific Software, you need to fundementally understand Software Engineering and best practices like design patterns, OO, Services, Scalability, Maintainability and all the other *-bilities.
I have run into way too many people in this field creating piles of crap software that end up getting foisted onto the next guy unlucky enough to role into town. Who, if he/she has no discipline end up creating the giant ball of mud.
My bottom line is that I care as much about my Software and doing it correctly as I do about my Science and always doing that to the best of my ability.
Brock.

Re: The Horrifying Future of Scientific Communication | Depth-First

Luke C. — Wed, 08 May 2013 13:51:48 -0000

While positioned slightly different than the examples you've presented, the Cambridge-based Journal of Visualized Experiments looks like an interesting take on the future of scientific publishing.

Re: Planes, Trains, and Organic Syntheses | Depth-First

Rich Apodaca — Sat, 04 May 2013 12:21:28 -0000

A group effort of some kind might actually be possible. I'm reminded of how CAS got its start - with 'volunteers'. Maybe not likely today, but at least possible.

Although volunteerism is well and good, I think the effort would have a much better chance of success if it could be linked directly in some way to participants' day-to-day activities.

Re: Planes, Trains, and Organic Syntheses | Depth-First

Rich Apodaca — Sat, 04 May 2013 12:16:23 -0000

I take it there is no standard (either XML-based, or JSON-LD-based) in cheminformatics for representing reaction networks. Has anyone thought about about creating one?

Not that I know of.

But I suspect that the challenges of creating even the most minimal graph simply linking molecules (nodes) with each other via an edge given a known transformation would far outweigh the challenges of representing the graph.

Re: Free Access to ACS Publications and Why You Can't Have It (Yet) | Depth-First

Rich Apodaca — Sat, 04 May 2013 10:47:37 -0000

May 4, 2013 - it seems to be fixed...

Re: Planes, Trains, and Organic Syntheses | Depth-First

Geoff Hutchison — Thu, 02 May 2013 15:02:55 -0000

If the ACS, RCS or other academic chemistry society started asking for reaction schemes in a chemical, readable format, like they do for crystal structures, crowdsourcing would not be needed. As it stands, we "destroy" such data by asking for pictures.

Re: My Favorite Eclipse Shortcut: Quick Fix

Leandro Guida — Thu, 25 Apr 2013 15:48:06 -0000

Or just press Ctrl + O which is the same thing.

Re: Planes, Trains, and Organic Syntheses | Depth-First

James — Thu, 25 Apr 2013 13:22:22 -0000

Re: accumulating data, this sounds like an opportunity for crowdsourcing from organic chemistry graduate students.

Re: Planes, Trains, and Organic Syntheses | Depth-First

Mark Fortner — Wed, 24 Apr 2013 10:37:09 -0000

Interesting article. I take it there is no standard (either XML-based, or JSON-LD-based) in cheminformatics for representing reaction networks. Has anyone thought about about creating one? You could use XGMML the native format for Cytoscape although you'd have to figure out how to represent the semantics of reactions. Also the arcs between the nodes in the diagrams above, don't really tell you how the chemist got from point A to point B. Something like "heat to 50C", "sprinkle in iron filings", etc.

Re: A Third Failed Test of ACS Articles on Request and How to Help | Depth-First

Rich Apodaca — Fri, 19 Apr 2013 10:44:47 -0000

Alain, your workaround appears to do the trick. Creating a free, alternate ACS id allowed me to view the AoR articles.

I'm still unable to view the AoR-linked articles through my membership-linked ACS ID. Odd.

Re: Free Access to ACS Publications and Why You Can't Have It (Yet) | Depth-First

Alain Borel — Thu, 18 Apr 2013 15:10:39 -0000

I can access the _mentioned_ J. Chem. Inf. Model. _paper_ as well (sorry for sending the post before re-reading it).

Re: Free Access to ACS Publications and Why You Can't Have It (Yet) | Depth-First

Alain Borel — Thu, 18 Apr 2013 15:08:37 -0000

I can access the J. Chem. Inf. Model as well. I have also verified that I am properly blocked if I try to access a random ACS paper through the regular channels (from Google Scholar, specificly).

Re: A Third Failed Test of ACS Articles on Request and How to Help | Depth-First

Alain Borel — Wed, 17 Apr 2013 15:12:46 -0000

Having no ACS ID, I had to create one. In order to do so, I had to provide an e-mail address. What I mean is that, just in case it should matter, I was careful not to use an e-mail address related to the university where I work (where we have an extensive ACS license): instead, I used one at yahoo.fr.

Re: A Third Failed Test of ACS Articles on Request and How to Help | Depth-First

Rich Apodaca — Wed, 17 Apr 2013 14:32:15 -0000

This is what I see (attached). I'm not sure why were getting two different results.

When you say "I used a private e-mail address", where did you use it and how?

Re: A Third Failed Test of ACS Articles on Request and How to Help | Depth-First

Alain Borel — Wed, 17 Apr 2013 14:03:01 -0000

No, I can download the fulltext in HTML or PDF without any problem.

Re: A Third Failed Test of ACS Articles on Request and How to Help | Depth-First

Rich Apodaca — Wed, 17 Apr 2013 11:22:38 -0000

Alain, your link now works, but generates a message on the page like the other AoR links I've tested:

"Your current credentials do not allow retrieval of the full text."

Do you also see this message? I tested on Firefox 20.0/OS X.

Re: A Third Failed Test of ACS Articles on Request and How to Help | Depth-First

Alain Borel — Tue, 16 Apr 2013 11:00:18 -0000

OK, I see... apparently my link uses an older syntax and the redirection doesn't work properly (it loses the + sign in the article code). Try this one instead:

pubs.acs.org/articlesonrequest...

As for the JPC C papers: I'm using Firefox 20 on OSX 10.6.8 in case the problem is browser- and/or OS-dependent.