Inference of Boolean models of genetic networks using monotonic time transformations

This paper considers the problem of inferring a Boolean network (BN) from gene expression data that is available as a frequently sampled time-series. The particular problems that arise are discussed and monotonic time transformations (MTT) are presented as a possible solution. Several different methods of clustering are used to form different transformations. The results with data generated by a simulation model show that the method presented can improve the inference performance in the described cases. The real-world measurements currently available are not yet suitable for testing the method because of the low sampling rates and the amount of noise present.