Van Cleve, Jeremy

Populations genetics of frequency dependent social evolution

Most mathematical models in the study of the evolution of social behavior (e.g., cooperation and conflict) make significant simplifying assumptions that make the analysis tractable. These assumptions, which can include weak natural selection and weak mutation rates, are often not met in nature and thus exclude important biological realism. Recent work in population genetics has made use of a mathematical tools from physics that allows some of these assumptions to be relaxed for non-social evolution. The aim of this work is to extend this work by using these tools to study social evolution and frequency dependence. Analyzing the accuracy of these new analytical results will require significant numerical simulations, which we will perform using a mixture of C and Python. Additional numerical simulations will be used in the case where analytical results are not accessible.

Software:

Tools (all free and/or open source)
• Python 2/3 with the following modules: numpy, scipy, matplotlib, h5py, cython, python-igraph
• R with the following libraries: igraph, ggplot2
• Libraries (at minimum): gsl, hdf5, igraph
• Compliers: gcc, g++, gfortran

• Julia

Staff:

• Jeremy Van Cleve (PI – UK)
• Elliott Greene

• Daniel Priego Espinosa, Added 01/29/2021

Grants:

None yet from the University of Kentucky HPC.

Evolution of regulatory networks for social traits We want to understand how the regulation of social traits evolves

By social traits we mean both social behaviors and socially relevant morphological and physiological traits. By regulation we mean both within-individual mechanisms such as gene regulatory and neural networks, as well as the regulation of behavior through social signals and responses. Our main research aim is to develop a quantitative theory that answers the following questions and test the theory using genomic, transcriptomic and behavioral data:
1. How are gene regulatory networks selected to produce and respond to social signals?
2. How does the architecture of regulation affect the evolution of social traits, in particular through producing social responses and indirect genetic effects (IGEs)
3. Where does selection act in the gene regulatory networks for social traits, and how does the regulatory network change during social evolution? In particular, to what extent is divergent social evolution driven by ”tinkering” with existing regulatory elements vs. adding new links and/or nodes to the regulatory network?

Software:

Tools (all free and/or open source)
• Python 2/3 with the following modules: numpy, scipy, matplotlib libraries, h5py, cython, python-igraph
• R with the following libraries: igraph, ggplot2
• Libraries (at minimum): gsl, hdf5, igraph
• Compliers: gcc, g++, gfortran

• Julia

Students and Staff:

• Jeremy Van Cleve (PI – UK)

• Elliott Greene

• Daniel Priego Espinosa

Collaborators:

• Erol Akçay (Collaborator – Asst. Prof. UPenn)
• Tim Linksvayer (Collaborator – Asst. Prof. UPenn)

Grants:

The evolution of phenotypic plasticity and bet-hedging

One important mechanism that buffers organisms against environmental variation is phenotypic plasticity. Responding to environmental cues is almost certainly costly, which suggests that plasticity only evolves when its benefits outweigh the costs. The degree of plasticity among organisms is large, and the level plasticity can vary according to a variety of variables. Though such diversity in plasticity immediately suggests diversity in costs, most models of the evolution of plasticity neglect the fact that costs might vary. I begin to address this problem by varying the cost of plasticity with either an ecological variable, growth rate, or a demographic variable, age. In the first set of analyses, I will build a analytical model (non-autonomous differential equation) and analyze the model numerically a mixture of C and Python. In the second set of analyses, I will build a quantitative genetic model analytically and simulation the results of the model using an individually-based model developed in C and Python.

Software:

Tools (all free and/or open source)
• Python 2/3 with the following modules: numpy, scipy, matplotlib, h5py, cython, python-igraph
• R with the following libraries: igraph, ggplot2
• Libraries (at minimum): gsl, hdf5, igraph
• Compliers: gcc, g++, gfortran

• Julia

Staff:

• Jeremy Van Cleve (PI – UK)

Grants:

None yet from the University of Kentucky HPC.

Publications: