Our software projects are all hosted on github
We have developed an extensive library for customizable forward simulations that is highly optimized to allow the simulation of large populations. We use it extensively in our daily research and extend it as need arises. It is simple, fast and flexible. We have published and application note in bioinformatics describing it. More in depth descriptions and examples can be found on the the FFPopSim site. The source code itself is hosted on github.
Estimating selection coefficients from time series data
Viral populations, rapidly developing cancers, or population in microbial evolution experiments change so quickly that we can observe how frequencies of different genotypes change over time. When many such mutations spread at the same time and interfere, estimating the underlying evolutionary model from observations is difficult. We have developed a method that fits a simple model of sequentially accumulating beneficial mutations to a series of genotype samples. The source code and documentation are available here.
We have developed a set of tools to decompose multiply labeled fluorescence microscopy images into contribution of the individuals dyes — a method known as spectral unmixing. Unmixing is easy when the spectra (mixing matrix) are known. We present methods to unmix image stacks when this mixing matrix is not or only partly known. Documentation and examples. The source code can be found here.