Annual Progress Report


Tomorrow will be meeting with my dissertation committee: Alan Hastings, Peter Wainwright, and Brian Moore. This provides a good opportunity to summarize the status of my current projects and my principle accomplishments for the year. Will try a tex’d document as I’ve done in previous years but here’s for some preliminary brainstorming.

  • Summary: In the past year, I’ve published one paper, given 4 talks (CPB, NIMBioS, Berkeley, IIASA), read 144+ articles and wrote 9000+ lines of code. I spent last summer at IIASA, passed my qualifying exam in January, and attended the Bodega Phylogenetics workshop in March. I’m working primarily on four projects. If my thesis had a unifying title, it would be something like: Inference in a noisy world

Evolution

Adaptive Dynamics

Comparative Phylogenetics

  • Labrid fin morphology paper
  • Quantifying the information in phylogenetic trees: robust model choice in comparative phylogenetics
  • logarithms and Brownian motion
  • Joint-probability solution?
  • Collection of smaller extensions to phylogenetic methods
  • Apply comparative methods in Primate brain morphology (with Daniel Mietchen)
  • Density dependence/equilibrium models in speciation rates. Model uncertainty in rates.

Ecology

Early Warning Signals

  • Misleading signals and more robust approaches.

Structured Population Dynamics (aka beetles)

  • Large effects of demographic noise through nonlinear dynamics in structured populations


### Talks & Conferences

  • Talks to: CPB (Davis), NIMBioS (Tennessee), Comp Bio students (Berkeley)
  • in-house talks to EDG (Fall & Spring), Hastings lab,
  • Courses: Anoles. Begun Monte Carlo. Phylogenetics Discussion Group (winter/spring).
  • Bodega Phylogenetics Workshop.
  • Lab group meetings: Hastings, Schreiber. occasionally Wainwright, Crutchfield,

Upcoming

  • Conferences this summer: Evolution & iEvoBio. CSGF. Open Science.
  • Arkin group at LBL

Methods and Misc

  • Technology/Science2.0 : Learned git for version management, github project hosting. Mendeley database for papers. OpenWetWare for lab notebooks. Forums.
  • Interviewed in Science Careers.
  • Help start UC_Davis_Open_Science group.
  • Grad Teaching Community.


#### Footnote on some stats

Thanks to the various Science 2.0 spirit technology I’ve employed in my research, it’s remarkably easy to get statistics on what I’ve done over the year. For instance, as I read articles I add them to my Mendeley database, making it easy to recover my article statistics, and also plot these trends over time. Similarly, version management software such as git makes it easy to track the progress of my codes. Will be interesting to revisit these stats next year. Meanwhile:

Articles: I’ve read 144 articles since November (when I started using Mendeley). Most read journal has been Evolution (25 articles), most read authors are Peter Wainwright and Johnathan Losos. 81 articles were in comparative methods. 6 have been reviews I’ve done for different journal submissions.

Code: I’ve written over 9000 lines of code in five projects: regime_shifts (1417), structured_pops (684), Comparative-Phylogenetics (2265), Adaptive-Dynamics (1326), wrightscape (3380).



Code Tricks ———–

  • Was curious how to quickly find the number of lines of code in my projects. While Github has several useful graphs, they haven’t yet added this feature. Fortunately this question was already answered on stackoverflow with a very nice little python script, which I’ve named gitlines and have added to my path. Running gitlines from anywhere in the project directories gives a list of total lines in each commit for the project.


Monday ——

  • 10a - 11:30a Steering committee meeting
  • Chat Sebastian re: linear noise approximation in patch dynamics, see earlier manuscript
  • 1p Sebastian lab meeting/pizza social, Discussed F1000, Kyle mentioned niche modeling paper in Science. Discussed which ToC’s people read, also compare to which they use or cite.
  • 5p Richard McElreath speaking to GTC on teaching mathematical modeling to biologists

Research

From discussion with Sebastian: Structured population dynamics need to consider stage transitions as simultaneous two-step events rather than two one-step events. results in slightly different form for evolution of the covariance matrix. Excepting this yesterday’s form is accurate.