Plenty of Fish in the Sea, or “Analysis & Results”

There are plenty of fish in the sea… Especially in the Green, Tennessee, Ohio, Cumberland, and Laurel river systems. I couldn’t go the whole summer without throwing that line in, could I?

971

Liz Lemon likes my joke. 

Update: I finished my pipetting. I finished gathering the genetic information at 25 loci for 169 individulas. (Or should I say fin-dividuals?) I finished sifting through those numbers to weed out the false alleles. In short, I FINISHED MY DATA COLLECTION.

I’m as happy as this guy. happy-fish

I’m left with three days for analysis before I leave New Haven. My computer and I have become fast friends. I have never considered myself a particularly tech savvy person, but lo and behold, I am writing rudimentary code. Along the way I’ve picked up the tools Tandem (which rounds nucleotide repeats) and STRUCTURE (which analyses genotype data to infer distinct populations).

My first round of analysis yielded five distinct populations. Each of the six sections in the picture below represents a locality that we sampled in the river system. Bar Plot K5

At first glance, it appears that populations 1, 2, the majority of 3, and 4 each form their own genetic clusters. A handful of specimen in population 3 (the dark blue amid the red) were mislabelled and will be taken out of the dataset.

Populations 5 and 6 are different from the others but similar to each other. However, when I ran only populations 5 and 6 together, I saw that…

d

…at a more specific level of analysis, they form their own genetic clusters as well!

I’m working with Professor Near and a postdoc in the lab, Rich, to continue this analysis. The next steps involve reorganizing the fin-dividuals in the dataset to reflect the geography of the river system. For instance, if population 4 represents the Upper Tennessee River, we will place those individuals at the top of the river system early in the dataset and those individuals at the bottom of the river system at the end of the dataset.

Further work will show if the microsatellite data supports the already established gene tree formed with mitochondrial data.