A new tool for fusioning PET and omics data. (1st February 2022)

• Record Type: Journal Article
• Authors: Povala, Guilherme; De Bastiani, Marco Antônio; Brum, Wagner Scheeren; Ferreira, Pamela C.L.; Bellaver, Bruna; Zatt, Bruno; Zimmer, Eduardo R.
• Is Part Of: Alzheimer's & dementia. Volume 17(2021)Supplement 4
• Page Start: n/a
• Page End: n/a
• Publication Date: 2022-02-01
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.052987
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: Positron emission tomography (PET) imaging plays a key role in the diagnosis of Alzheimer's disease (AD). The idea of integrating PET and omics data can provide blood‐new insights about AD pathophysiology. In this context, state‐of‐the‐art techniques such as transcriptomics hold much potential for discovering altered biological processes in AD and other neurodegenerative diseases. However, tools for integrating PET and omics data are still unexplored. Here, we developed a new method that combines blood transcriptomics with PET data. We hypothesized that our method will allow advancing our understanding of AD neurobiology and potentially unravel clinically relevant new peripheral biomarkers. Method: Imaging and transcriptomics data were acquired from the Alzheimer's Disease Neuroimaging Initiative (ADNI). Microarray gene expression profiling from blood samples of 69 cognitively unimpaired (CU) and 158 mild cognitively impaired (MCI) individuals were submitted to differential expression (DE) analysis using the limma R package. Genes obtained from DE analysis were selected to undergo integration with [ 18 F]Fluorodeoxyglucose (FDG)‐PET images using voxel‐wise generalized linear regression (GLR) models adjusted for age, gender and APOEε4 (RMINC package). Result: The DE analysis resulted in 1232 differentially expressed genes (DEGs) in CU vs MCI individuals (p‐value<0.05). For each DEG, the GLR computed the voxel‐wise association between FDG‐PET and gene expression, resulting in t‐value maps. Afterward, only gray matter voxels presenting absolute t‐values higher than 2.3 were preserved. Then, the volumes of interest (VOI) presenting at least 30% of voxels with a significant correlation obtained from the t‐value map were filtered, which resulted in 268 significantly correlated genes throughout the brain. Next, a map of DEG counts in each VOI (Figure 1a) and maps with the proportion of up‐ (Figure 1b) and downregulated DEGs (Figure 1c) were created. An overall map of altered VOIs is displayed in Figure 1d. Finally, we identified that the putamen was the VOI with the most significantly correlated DEGs. Conclusion: Here we presented a pipeline for integrating PET neuroimaging with transcriptomics data. This method allows for the identification of potential novel biomarkers and altered peripheral biological pathways.