Inferring Better Gene Regulation Networks from Single-Cell Data
Networks can provide a graphical representation of complex systems, including gene regulation processes. They can also provide a basis for further quantitative and computational analysis and modelling of biological systems. Single-cell technology is now allowing us to probe molecular mechanisms and processes at unprecedented detail, and there is hope that we can learn better, more detailed network models from such data, to help us understand the mechanisms underlying cellular processes. However, learning the structure of networks is notoriously difficult for at least two reasons: (i) network inference is a statistically challenging problem and (ii) the naive picture of static networks may be fundamentally inapplicable to the description of biological systems. Here, I will give some overview of the basic problem, discuss a set of promising network inference methods and how to validate them and outline how we can go beyond the limitations imposed by static network models.