Blue skies come to earth

I was up mad early yesterday morning and fielded an e-mail from my old gaffer, sent late the previous night, looking for a recent reference for one of her brighter students, who was asking some on-the-button questions that predicts a bright and shining future in science. Looking to me, with my two week event horizon, for that sort of thing is a bit of an ask but I can still do research so I went looking. It turned out to be a paper that we had talked about a couple of years ago. It showed that the work I had pursued twenty years ago as quintessential pure science which had no earthly 'use' in a curing-cancer or feeding-the-millions sort of way, turned out after a lapse of two decades to be really important in biotechnology.  Score 1 for funding blue-skies research. Then, as I left for work, I heard about a letter signed by 900 Irish Scientists which urged the government to start funding less obvious projects and open a seam of support for pure, blue-skies, higher risk research. I worked myself into a lather about that because I felt that I had been disrespected by not being asked to sign too.  But it was for sure a blue-sky day.

When we came back to Ireland in 1990, I secured a job doing a particular/peculiar class of research involving the analysis of DNA and protein sequences. You may like to have a quick refresher of how the genetic code works. It was interesting and intellectually challenging and hinged on the fact that genes of low-to-mid-range expression used 'codons' (the DNA triplets that code for particular amino acids in proteins) more or less indiscriminately while highly expressed genes use a biased subset of codons and you can use this to infer the expression level of an unknown gene. Almost all microbes had such an expression pattern but each species used its own characteristic subset of codons in its highly expressed genes. It was canny enough in its day, the data was patchy before mega-genome sequencing kicked in ten years later. You had to work hard and read a lot to make sense of the data and find the signal in amongst the noise. But we helped set the ground-rules for the field of synonymous codon usage research. I've had three good ideas in 40 years of science, and one of them came out of that project. I'm not making any such bigging-me-up claims for my 30-years-ago research on domestic cat genetics that came back to the spotlight for 15 minutes on Tuesday.

I hasten to say that I was too thick to appreciate at the time how this codon useless usage knowledge could be mobilised for anything more than honing my crap-detector and generating some papers that were cited a few dozen to about 200 times by other scientists. The average paper is cited less than once and 45% of all papers are published and never referred to again . . . by anyone: so we had some modest impact on science. But in October 2013, a paper was published in Nature showing that if you want to make some useful product (insulin, say, or in this case an enzyme that will efficiently degrade chitin for better health) using current biotechnological techniques you'd better pay attention to codon usage.  In particular you should modify your gene of interest so that it looks as much as possible like a highly expressed gene of the carrier species that will be grown in a 10,000 lt vat and generate kilos of your Amazing Therapeutic Protein. If you just lash in the human gene, as it is found in nature, it won't get fast-tracked by the expression vector, and may not work at all. As I say, I didn't have any of the key insights that got this paper published in Europe's premier science journal, but it is fair to say that they stood on the shoulders of our 'useless' research from the end of the last century.  I might say "I have done the state some service", as I prepare to exit stage right.