Efficient and Efficacious Cellular Reprogramming with Edgetic Perturbations (EdgeCR)

The project at a glance

Start date:

11 Jan 2024
Duration in months:

48
Funding:

National Science Centre, Poland / Luxembourg National Research Fund
Principal Investigator(s):

Jun PANG

��ͷ��

Complex diseases such as Parkinson’s are not typically caused by mutations in single genes, but rather by disruptions across interconnected networks of genes and their products, known as gene regulatory networks (GRNs).The complexity of GRNs coupled with the high costs and time consuming nature of wetlab experiments preclude exhaustive trial-and-error wetlab analyses. Consequently, there is a growing need for computational approaches to predict promising strategies for cellular reprogramming— techniques that can transform one specialised cell type into another in order to repair or replace damaged tissue and restore healthy function. The EdgeCR project addresses that need by developing a computational framework designed to identify effective cellular reprogramming strategies by modelling and manipulating GRNs. The project will build Boolean network (BN) models from gene expression data. BNs represent the activity of biological entities as active or inactive. This allows for reasoning about causal relationships between entities with few parameters. In BNs, cellular reprogramming is usually modelled by setting individual genes to active or inactive. Moving beyond classical methods that only target individual genes (nodes), EdgeCR will also consider “edgetic” perturbations—refined interventions that directly modify the interactions (edges) between genes, offering more nuanced control over cellular behaviour. Novel algorithms, enhanced by deep reinforcement learning, will allow the efficient exploration of both gene- and edge-level interventions, even in the vast search spaces of large GRNs. EdgeCR will apply its computational predictions to the reprogramming of astrocytes into dopaminergic neurones, addressing cellular loss found in Parkinson’s disease. By integrating AI-driven, in silico analysis with biological validation, the project aims to deliver a scalable, state-of-the-art toolkit for guiding and controlling cellular reprogramming through detailed, actionable insights into both the nodes and edges of gene regulatory networks.