Saturday 15 November 2014

Cites seeing

I've had a bumbly sort of career in science (so far! it's not over till the fat Blobby sings).  Not stellar but not without faults.  From my perspective, the major problem has been 'finish', I've had a few (3) good ideas, I've worked really hard to develop projects but I've often-and-often fallen at the last hurdle: having solved the problem to my own satisfaction, I go off the boil and move on to the Next Big Thing rather than sitting down, writing it all up and getting it published.  I wrote a while back about how the landscape of publication is skewed by the process of science - there are far too many papers adding warts to the portrait of some Important Idea while other equally important ideas are barely sketched out.

But there's no point in carrying out some research, however diligently, if nobody reads or cares about the result. That's a waste of some graduate student's time and intellectual energy and a waste of tax-payers' money.  [Except insofar as the student gotten some training in benchwork, statistics, critical thinking, interpersonal relationships, or ExCel].  How do we know which of the papers are wasting space?  We ask whether some other scientist has referenced the work in a paper of their own.  Such an acknowledgement is called a citation and 50 years ago last month Eugene Garfield started to collect, index and distribute these citations.  Nature, Europe's scientific journal of reference, has just reviewed progress in the field. And produced its own mountain of data.  According to their records, 57 million scientific papers have been published since the beginning of time, millions since WWII, rather fewer at the beginning when Cuvier, Darwin, Kepler and Newton started working on the foundations. 45% of those papers (25.3 million) have sunk without trace, nobody else having found a single thing of value in the effort - heck, not even the authors of the paper in their own subsequent work.  The only positive thing about this sorry state of affairs is that 'papers' are increasingly published only on-line, so only electrons are being wasted rather than trees.  You can look up your scientific cousin (use lastname initials) to see how big a cheese she is.  My own contribution has been bumbly enough with some reasonably referenced material: my 'top' paper, on the nematode worm Caenorhabditis elegans, having garnered 253 citations.The rest of my opus dribbles off with 198, 197, 184, 136, 115, 109, 103, 97, 92 . . . 32, 29, 23, 19, 10, 10, 7, 7, 5, 5 and then disappears into the grass.  If you tally it all up, my citation count is about 2300: mid-range, not entirely worthless, not very exciting.

This last week at The Institute, we've been going great guns for Science Week. As part of this jamboree I invited a far more successful scientist down from Dublin to talk to our students.  Des Higgins, who has had a walk-on part on The Blob before, was asked to talk about The Joy of Molecular Evolution. The title implied that what we both do is as exciting as ...Sex ...Cooking ...the Lord is My Strength or ...Painting which is the order Google offers for completion of The Joy of...  I asked Des a couple of months ago and in the interim Nature published its analysis of exciteing papers (geddit?). This revealed that Professor Higgins is by far and away Ireland's most cited scientist! This is almost all on the back of the discovery and development of a handy and effective method of aligning biological sequences.  Since he had the kaCHING! idea in 1987, he has written a program for carrying out this key infrastructural task and over the last 20+ years developed it to crest the tsunami of sequence data that has flooded our world. Each development required an additional paper to explain how it worked and how the new material was implemented.  The most useful of these ClustalW was published 20 years ago next week on 22 Nov 1994. In the intervening two decades it has gathered an astonishing 48000 citations - on its own dwarfing by 20x my lifetime's work and making it the 10th most cited paper of all time. Reports on other stages in the development - Clustal1-4, ClustalV, ClustalX, ClustalΩ - have boosted this total to something more that 115,000 cites for this one Big Idea.

Hats off, it's a huge achievement indicating a HUGE contribution to the field . . . and also an indication of how huge the field of bioinformatics and molecular evolution has become. But, as Des pointed out in his talk, citations accrue more to successful techniques than to paradigm shifting ideas. Crick and Watson's description of the structure of DNA, which started this whole gallop of in 1953, has only been cited a tad over 10,000 times. Nirenberg and Matthaei [Blob-trib], who started working out how the genetic code worked in 1961 have been acknowledged only 1700 times.  Now, if only more people would read The Blob, I wouldn't feel I was wasting my time . . . not to mention all the electrons.

No comments:

Post a Comment