Candida albicans is the most common cause of life-threatening disseminated fungalinfections. It is capable of undergoing phenotypic switching between two distinct cell
types, named ‘white’ and ‘opaque’. The white-opaque switch is controlled by a complex
genetic regulatory network (GRN) that consists of transcriptional regulators (TRs). The
advent of methodologies for profiling mRNA transcript levels and specific protein-DNA
interactions at a genome-wide level has greatly expanded our ability to determine the
structure and output of genetic regulatory networks, however uncovering the logic of how
these networks function remains a challenging endeavor. The field of genetic regulatory
network inference aims to meet this challenge by using computational modeling to derive
the structure and logic of GRNs based on the experimental data provided by these genome-wide
approaches. Boolean, probabilistic, ODE-based, and other models have been developed
to infer GRNs. However, most existing models do not incorporate dynamic transcriptional
data, since it has historically been less widely available in comparison to “static” transcriptional
data.
In this work, a novel evolutionary algorithm-based ODE model that integrates kinetic
data and considers “static” transcriptional profiles as attractors has been developed to infer
GRN structure. The model performed well on both in-silico and real-life datasets, and it
was able to predict unknown transcription profiles produced upon genetic perturbations of
the Candida albicans white-opaque GRN. Those genetic perturbations can be engineered
in vivo and the result can be utilized to either support or further refine the model. Therefore,
the model allows for an iterative refinement strategy to decipher GRN and verify its reliability:
the model facilitates GRN candidate selection for experimentation and the experimental
result in turn provides validation or improvement for the model.