You will very rarely see me reading research on Drosophila melanogaster for the sole reason that fruit flies creep me out. Close up, they’re also ugly and in real life, they’re annoying. However, an association study between SNPs and sleeping time is just too interesting to pass up.

Briefly, MacKay’s lab measured sleep time in several inbred Drosophila lines by measuring movement in a glass tube. Whenever lack of movement was recorded for 5 minutes or more, they assumed that the fruit fly was sleeping. They recorded this data, sequenced the genome of each cell line, rearranged both sets of data on a spreadsheet and came up with preliminary data to an associative study. Now, while their methodology seems pretty arbitrary to me (do fruit flies really fall asleep in 5 min? Or are they merely dozing/resting? Do they twitch in their sleep?), I am not a Drosophila expert, nor am I a sleep scientist.

To me, the most interesting thing about this study weren’t the genes re-identified as having a role in sleep, nor the massive amount of time and effort required to conduct this study. The most interesting thing I found was a suggestion from Altshuler, who says that sleeping time “could arise from a combination of many variants, each with a tiny effect, or from interactions between several genes”.

“…complex traits can’t be explained by single genes with big effect.”

It’s taken us a long time to appreciate such a simple, logical concept. Very few things in genetics have a binary result. You’d think Gene A turns on or Gene A can turn off, giving phenotype versus no phenotype, but Mendelian genetics has lied to us here. This on:off, yes:no relationship may work for measures of very simple and basic phenotypes such as Rh- and Rh+, ABO blood type, albino versus pigmented skin (but note that I did not state the differences in skin coloring or tone) and eye color – basically for nominal categorizations of a phenotype.

1 Gene, 1 Phenotype will never hold true for 99% of everything else in nature that basically follows a continuous data spectrum. Life is simply too complex for every specific characteristic to be determined by 1 specific gene. We only have so much space to store our genetic information.

But… if the scientific community is still struggling to comprehend the complexities of gene interactions in the production of a single phenotype, how much worse is our ability to convey that to the lay public. After all, guess what we’re teaching our undergraduate science students in universities?