Imaginings of a Livestock Geneticist

I'm Jeremy Howard, and I am trained as a quantitative livestock geneticist. When I started my master's program, I was curious about how to code up models used in routine livestock evaluations and make them computationally efficient. Like many animal breeding students starting out, I had little to no knowledge of coding. As I progressed, I became more comfortable with SAS, then R, and eventually jumped headfirst into C++ during my PhD, where I developed a livestock simulation tool. Over the years, I accumulated a number of functions commonly used in livestock evaluations, including methods to create pedigree and genomic relationship matrices. In addition, I expanded my work to include machine learning and optimization functions. The functions are described in the 'Code' link, and in some cases, software was developed. The C++ code is always a work in progress, aiming to make it faster and minimize memory requirements. My hope is that this provides a good starting point for those wanting to learn about generating breeding values, relationship matrices, machine learning models, or optimization functions. I hope you find this useful in your endeavors and that it gets you excited about coding up new projects.