Integration of Imaging (epi)Genomics Data for the Study of Schizophrenia Using Group Sparse Joint Nonnegative Matrix...

Area of Science:

• Neuroscience
• Genetics
• Bioinformatics

Background:

• Schizophrenia (SZ) is a complex disorder with multifactorial etiology.
• Single nucleotide polymorphisms (SNPs), functional magnetic resonance imaging (fMRI), and DNA methylation are key biomarkers for SZ research.
• Integrating these diverse data types presents a significant challenge.

Purpose of the Study:

• To develop a novel computational model for integrating SNP, fMRI, and DNA methylation data in SZ.
• To identify multi-dimensional molecular and neural modules associated with SZ.
• To explore the regulatory mechanisms underlying SZ at multiple biological levels.

Main Methods:

• A group sparse joint nonnegative matrix factorization (GSJNMF) model was proposed.
• The model projects heterogeneous data onto a common feature space.
• Group structure information from each dataset was incorporated.

Main Results:

• The GSJNMF model successfully integrated SNP, fMRI, and DNA methylation data.
• Identified multi-dimensional modules showed significant correlations between genomic factors and brain activity.
• Applied to real SZ data from the MCIC, the method identified significant biomarkers.

Conclusions:

• The developed GSJNMF model effectively integrates multi-modal data for SZ research.
• Identified biomarkers facilitate the discovery of SZ-associated genes and brain regions.
• This integrative approach offers new insights into the complex regulatory mechanisms of schizophrenia.