Knowledge-based probabilistic representations of branching ratios in chemical networks

Abstract

We present the use of Nested Dirichlet distributions to represent uncertain branching ratios in chemical networks. The interest is twofold: (1) to preserve the structure of experimental data by imposing sum-to-one representations; and (2) to be able to introduce totally unknown subsets of branching ratios (missing data). These points are central to sound uncertainty propagation and sensitivity analysis in complex chemical networks.