Gene Regulatory Networks (GReNes) are involved in cell functions and analysis of pathways. GReNes are also useful for inferring the relationships for analyzing and inference there is need of development of good Gene Regulatory Networks (GReNes) approach or models. These models include estimation of molecular interactions, design of simulation systems, estimation of system irregularities via perturbation analysis, biomedical treatment estimation, drug design simulations, etc. Design of GReNes is a complex process, and involves modelling of operations like correlation estimation, activation function design, fuzzy cascade of signals, etc. In order to perform these operations various GReNes are proposed, which include single cell Graph Neural Networks (scGNN), boosting GReNes, ensemble trees based GReNes, reverse engineered Bayesian model based GReNes, and single-cell regulatory network inference and clustering (SCENIC). Each of these models has scalability issues, which limits their deployment capabilities to application specific systems. In order to remove this drawback, the underlying paper proposes an augmented GReNe that uses multiple activations via cascading simpler networks. Due to augmentation and multiple activations, accuracy of the proposed model is observed to be 8% better when compared with state-of-the art models when inferred from single cell transcripts.

Augmentation, cascade, Gene Regulatory Network, multiple activation, scalability