Decoding neuronal gene expression: integrative insights from omics and AI

Area of Science:

• Computational neuroscience
• Neuroinformatics
• Systems biology

Background:

• Neuronal function and disease vulnerability are controlled by complex regulatory programs.
• High-throughput omics and neuroimaging offer detailed data but bridging molecular and macro-scale brain information is an informatics challenge.

Purpose of the Study:

• To review the evolution of computational frameworks in neuro-omics.
• To evaluate these frameworks in the context of Alzheimer's disease, schizophrenia, and epilepsy.
• To identify key challenges in neuro-informatics for cross-scale brain analysis.

Main Methods:

• Synthesis of computational frameworks in neuro-omics.
• Evaluation of methodologies including co-expression modules, graph neural networks, and foundation models.
• Case study analysis within Alzheimer's disease, schizophrenia, and epilepsy.

Main Results:

• Computational frameworks have advanced from descriptive co-expression modules to causal graph neural networks and cross-scale foundation models.
• Methodologies are being applied to understand complex neurological disorders.
• Critical bottlenecks exist in data harmonization, spatial alignment, and causal interpretability.

Conclusions:

• The evolution of computational tools is crucial for integrating multi-modal neuroscientific data.
• Addressing informatics challenges is key to advancing our understanding of brain disorders.
• Future work should focus on improving data integration and causal inference across scales.