Wednesday notes

ApGenome

Aphaenogaster sample Genomic-tip DNA extraction day 2.

Nandrop results terrible!!! No idea why - saw beautiful strands of DNA at isopropanol precipitation stage and followed exact same protocol as previous extraction with Pogo after that.

Re-precipitated sample (500 ul) with 750 ul 70% EtOH. Centrifuged 10 min top speed. Air-dry. Re-hydrate in 200 ul Buffer AE and Nanodrop in morning.
I had saved Genomic-tip for this event. Added 1 ml Buffer QF. Precipitated in EtOH and centrifuged. Air-dry. Re-hydrate in 200 ul Buffer AE and Nanodrop in morning.

Website

While setting up jekyll for static website on xubuntu laptop, decided to write quick notes on install process:

Install jekyll, requires an up-to-date version of ruby

sudo apt-get install ruby1.9.1-dev
sudo gem install jekyll
# Need `pandoc-ruby` gem for pandoc compatability
sudo gem install pandoc-ruby

That’s it!

Clone the website and build or serve locally!

git clone https://github.com/johnstantongeddes/johnstantongeddes.org.git jekyll serve

Of course, this is for an install of my own website…a new website would require some more steps.

Added javascript to base html layout to allow mathjax in posts.

Reading

De Mita, S., Thuillet, A.-C., Gay, L., Ahmadi, N., Manel, S., Ronfort, J. & Vigouroux, Y. (2013). Detecting selection along environmental gradients: analysis of eight methods and their effectiveness for outbreeding and selfing populations. Molecular Ecology, 22, 1383–1399.

challenging to detect loci under selection using molecular data
authors evaluate 3 correlation-based and 5 differentiation-based methods using simulated data with different sampling schemes and migration patterns
all methods perform well when selection is strong
logistic-regression based methods haver higher false-positive rates
correlation-based methods have greater power when selection is weak
CWDRP (Coop et al 2010 Genetics) method has best power when selection is weak with low false positive rate
deviation from island model (e.g. isolation by distance) increases false positive rate for logistic regression based methods
selfing reduces power for differentiation-based methods and increases false positive rate for logistic-regression methods
sampling more than one individual per population increases false positives in logistic regression. Differentiation-based methods (especially FLK) improved with more samples per pop
CWDRP method computationally-slow and requires population-level frequency data
differentiation-based methods require a large number of populations (48) and 8-10 samples per pop
more thoughts on the implications of this paper for our Medicago work in next post