Skip to Main Content
Considerable attempts have been made to develop models and learning strategies to infer gene networks starting from single connections. However, due to noise and other difficulties that arise from making measurements at the meso and nano levels, these so called bottom-up approaches have not been of much success. The need for methods that use a top-down approach to extract global statistics from expression data has emerged to deal with such difficulties. This paper presents a theoretical framework that employs global statistics learnt from gene expression data to infer different network structural properties of large- scale gene regulatory networks. The framework is inspired by genetic algorithms and designed with the aim to address the different weaknesses in existing approaches. Experimental results show that the developed system is more superior to previously published results.