Linking Gene Sequences to Brain Connectivity Alterations in Schizophrenia

Abstract

Schizophrenia is a complex neuropsychiatric disorder characterized by altered brain connectivity patterns and strong genetic underpinnings. This study develops an intelligent computational system that integrates genomic data with functional magnetic resonance imaging to establish meaningful correlations between genetic variations in key schizophrenia-associated genes (DISC1, LRRTM1, DRD2) and altered brain connectivity patterns. The proposed methodology uses gene expression analysis combined with graph-based brain network analysis from fMRI data. Feature extraction involves constructing connectivity matrices and computing topological network measures such as clustering coefficients and global efficiency, while machine learning algorithms are trained to classify schizophrenia patients and identify at-risk individuals using integrated genomic and neuroimaging features. This research expects to deliver a comprehensive gene-brain analysis system capable of predicting schizophrenia risk based on these integrated profiles. It aims to identify novel biomarkers linking specific genetic variants to connectivity alterations in brain regions such as the prefrontal cortex and hippocampus. Anticipated outcomes include the development of interpretable visualizations, like brain maps and connectivity networks, offering insights into how key genes impact neural function. Ultimately, this interdisciplinary approach is designed to advance precision psychiatry by supporting earlier diagnosis and enabling more personalized treatment strategies.

Introduction

This project focuses on understanding the relationship between genetic variations and brain connectivity changes in schizophrenia by integrating genomic data (gene expression and SNPs) with functional MRI (fMRI) data. It investigates schizophrenia-associated genes such as DISC1, LRRTM1, and DRD2 to determine how genetic mutations influence structural and functional brain networks. The study combines bioinformatics, neuroscience, and machine learning to identify biomarkers that can support early diagnosis, personalized treatment, and precision psychiatry.

The research addresses the limitation of traditional studies that analyze genetics and brain imaging separately by developing a multimodal framework. Its objectives are to identify genetic biomarkers, model abnormal brain connectivity, correlate genetic variants with brain changes, and build predictive machine learning models for schizophrenia classification and risk prediction.

The project includes preprocessing genomic and fMRI data, dimensionality reduction using PCA, feature fusion, and classification using Logistic Regression, Random Forest, and XGBoost. Model performance is evaluated using accuracy, precision, recall, F1-score, and AUC, while SHAP is used to explain model predictions and identify important gene–brain relationships.

The literature survey highlights previous work in imaging genomics, graph theory, machine learning, and multimodal data integration, while identifying gaps such as limited causal analysis, small datasets, lack of longitudinal studies, and insufficient integration of genetics with brain connectivity.

The research methodology follows a structured pipeline consisting of data acquisition, preprocessing, feature extraction, multimodal fusion, model training, explainability, visualization, and patient-level prediction. The project uses 10,000 samples of gene expression, fMRI connectivity, and SNP data stored in NumPy arrays. Software tools include Python, Scikit-learn, XGBoost, SHAP, NumPy, Pandas, Matplotlib, and Seaborn.

Conclusion

This project successfully demonstrates the development of an intelligent predictive framework that integrates gene expression data and fMRI-based brain connectivity features to identify potential biomarkers for schizophrenia. By employing advanced data preprocessing, feature selection, and fusion techniques, the system effectively combines genetic and neuroimaging information to enhance understanding of the disorder’s underlying biological mechanisms. Machine learning algorithms such as Logistic Regression, Random Forest, and XGBoost were applied to the fused dataset to classify schizophrenia patients and healthy controls with high accuracy. The integration of computational genomics and neuroimaging analysis provides a novel perspective on how genetic variations (e.g., in DISC1, LRRTM1, DRD2) may influence abnormal brain connectivity patterns observed in schizophrenia. The framework not only facilitates early detection and diagnosis but also contributes to the growing field of precision psychiatry, promoting data-driven approaches for mental health research. Furthermore, the project’s emphasis on interpretability enables visualization of gene–brain relationships, supporting biological validation and clinical insights.

References

[1] Q. Ji and Y. Song, \"Schizophrenia-associated alterations in centrality patterns of functional brain networks,\" Scientific Reports, 2024. DOI: 10.1038/s41598-024-54024-5 [2] V. Carceller-Parramon and D. Mastroeni, \"Sex differences in gene expression in schizophrenia: A meta analysis,\" Molecular Psychiatry, 2024. DOI: 10.1038/s41380-023-02206-w [3] Y. Liu, H.-Q. Qu, X. Chang, et al., \"Expansion of Schizophrenia Gene Network Knowledge Using Machine Learning Selected Signals From Dorsolateral Prefrontal Cortex and Amygdala RNA-seq Data,\" Frontiers in Psychiatry, vol. 14, 2023. DOI: 10.3389/fpsyt.2023.1123942 [4] A. Hudon, M. Beaudoin, K. Phraxayavong, et al., \"Exploring the Intersection of Schizophrenia, Machine Learning, and Genomics: Scoping Review,\" Frontiers in Genetics, vol. 14, 2023. DOI: 10.3389/fgene.2023.1128072 [5] N. Lu, L. Tian, V. D. Calhoun, et al., \"Brain function, structure and genomic data are linked but show different sensitivity to duration of illness and disease stage in schizophrenia,\" Molecular Psychiatry, vol. 27, pp. 1551–1561, 2022. DOI: 10.1038/s41380-021-01188-9 [6] G. Chand and V. K. Potluri, \"Network analysis reveals abnormal brain connectivity in schizophrenia,\" Scientific Reports, vol. 10, 2020. DOI: 10.1038/s41598-020-69271-z [7] J. Zhang and S. Morgan, \"Generative models identify mechanisms of altered connectivity in schizophrenia,\" NeuroImage, vol. 201, 2019. DOI: 10.1016/j.neuroimage.2019.116103 [8] X. Zhang, \"Generative network models of altered structural brain connectivity in schizophrenia,\" Schizophrenia Bulletin, vol. 45, no. 4, pp. 804–815, 2019. DOI: 10.1093/schbul/sbz067 [9] K. C. Skåtun and R. J. Dolan, \"Global brain connectivity in schizophrenia and bipolar disorder,\" Human Brain Mapping, vol. 37, no. 3, pp. 873–885, 2016. DOI: 10.1002/hbm.23052 [10] Y. He and A. Evans, \"Altered Small-World Brain Networks in Schizophrenia during Working Memory Tasks,\" Cerebral Cortex, vol. 22, no. 2, pp. 372–380, 2012. DOI: 10.1093/cercor/bhr300 [11] S. Yao et al., “Connecting genomic results for psychiatric disorders to human brain cell types and regions reveals convergence with functional connectivity,” Nature Communications, vol. 16, no. 395, 2025. doi: 10.1038/s41467-024-55611-1. [12] J. Chen et al., “A schizophrenia-related genetic–brain–cognition pathway revealed in a large Chinese population,” Schizophrenia Bulletin, vol. 47, no. 3, pp. 710–721, 2021, doi: 10.1093/schbul/sbaa141. [13] N. S. Shaik, T. K. Cherukuri, V. D. Calhoun, and D. H. Ye, “Multi-modal Imaging Genomics Transformer: Attentive Integration of Imaging with Genomic Biomarkers for Schizophrenia Classification,” arXiv preprint arXiv:2407.19385, Jul. 2024. [14] J. Xiang, C. Ma, X. Chen, and C. Cheng, “Investigating connectivity gradients in schizophrenia: Integrating functional, structural, and genetic perspectives,” Brain Sciences, vol. 15, no. 2, p. 179, Feb. 2025, doi: 10.3390/brainsci15020179. [15] S. Xia et al., “Functional connectivity signatures of schizophrenia revealed by neuroimaging and transcriptomic data,” Frontiers in Psychiatry, vol. 12, p. 147, 2021, doi: 10.3389/fpsyt.2021.602227. [16] C. N. Gupta et al., “Genetic markers of white matter integrity in schizophrenia revealed by parallel ICA,” Frontiers in Human Neuroscience, vol. 9, no. 100, pp. 1–14, Mar. 2015, doi: 10.3389/fnhum.2015.00100. [17] J. Liu et al., “Combining fMRI and SNP data to investigate connections between brain function and genetics using parallel ICA,” Human Brain Mapping, vol. 30, no. 1, pp. 241–255, 2009, doi: 10.1002/hbm.20508. [18] K. M. Anderson et al., “Transcriptional and imaging-genetic association of cortical interneurons, brain function, and schizophrenia risk,” Nature Communications, vol. 11, no. 2889, pp. 1–14, 2020, doi: 10.1038/s41467-020-16710-x. [19] S. Ghosal et al., “G-MIND: An end-to-end multimodal imaging-genetics framework for biomarker identification and disease classification,” arXiv preprint arXiv:2101.11656, Jan. 2021.

Copyright

Copyright © 2026 Aditya Pujari, Aditya Jagtap, Aditi Dixit, Prof. Anup Dange. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.