Biomedicine is the science of networks. In fact, many biological systems such as protein contact networks, protein-protein interaction networks, metabolic pathways gene regulatory networks, just to name a few, have a ‘natural’ network-based representation.
The extraction of useful information from biological networks has been an active research topic since the last two decades and the explosion of interest in modelling biological processes has found widespread acclaim due to the advent of automatic learning systems driven by machine learning and artificial intelligence. Notwithstanding their great potential, the vast majority of the learning systems are 'black boxes': that is, the end-user cannot understand why and how the learning machine made its decision. Conversely, different style of analyses may aim at reaching the biological meaningfulness of the results rather than focusing only on maximizing the predictive performance. This different style of analysis stems from the observation that mathematical, technical and computational knowledge must give relevant and meaningful scientific hints to the field-experts - i.e., biologists, chemists, physicians and the like - and cannot be considered as a mere predictive tool.
In this Research Topic, we aim at collecting research articles merging statistical and structural network-based modelling of biological systems with an emphasis on the model explainability. Specifically, we welcome both methodological papers, proposing novel techniques for explainable network inference and analysis, and application papers, showing said techniques in action for solving real-world problems.
Areas of interest to this Research Topic include, but are not limited to:
• Design of novel predictive methods over networks
• Explainable artificial intelligence
• Analysis and modelling of large-scale networks in biomedicine
• Hypergraph modelling of complex systems
• Structure-Activity Relationships in drug discovery and predictive toxicology
• Biomolecules structural and functional analysis
• Physiological signal and image analysis by network-based approaches
• Clinical Diagnosis by the exploitation of symptom constellation network
We strongly encourage the submission of Original Research and Methods articles, but also welcome submissions in the format of Data Report, Perspective, Systematic Review, Review, and Mini Review.
Biomedicine is the science of networks. In fact, many biological systems such as protein contact networks, protein-protein interaction networks, metabolic pathways gene regulatory networks, just to name a few, have a ‘natural’ network-based representation.
The extraction of useful information from biological networks has been an active research topic since the last two decades and the explosion of interest in modelling biological processes has found widespread acclaim due to the advent of automatic learning systems driven by machine learning and artificial intelligence. Notwithstanding their great potential, the vast majority of the learning systems are 'black boxes': that is, the end-user cannot understand why and how the learning machine made its decision. Conversely, different style of analyses may aim at reaching the biological meaningfulness of the results rather than focusing only on maximizing the predictive performance. This different style of analysis stems from the observation that mathematical, technical and computational knowledge must give relevant and meaningful scientific hints to the field-experts - i.e., biologists, chemists, physicians and the like - and cannot be considered as a mere predictive tool.
In this Research Topic, we aim at collecting research articles merging statistical and structural network-based modelling of biological systems with an emphasis on the model explainability. Specifically, we welcome both methodological papers, proposing novel techniques for explainable network inference and analysis, and application papers, showing said techniques in action for solving real-world problems.
Areas of interest to this Research Topic include, but are not limited to:
• Design of novel predictive methods over networks
• Explainable artificial intelligence
• Analysis and modelling of large-scale networks in biomedicine
• Hypergraph modelling of complex systems
• Structure-Activity Relationships in drug discovery and predictive toxicology
• Biomolecules structural and functional analysis
• Physiological signal and image analysis by network-based approaches
• Clinical Diagnosis by the exploitation of symptom constellation network
We strongly encourage the submission of Original Research and Methods articles, but also welcome submissions in the format of Data Report, Perspective, Systematic Review, Review, and Mini Review.