FDR-Corrected Sparse Canonical Correlation Analysis With Applications to Imaging Genomics

Abstract

Reducing the number of false discoveries is presently one of the most pressing issues in the life sciences. It is of especially great importance for many applications in neuroimaging and genomics, where data sets are typically high-dimensional, which means that the number of explanatory variables exceeds the sample size. The false discovery rate (FDR) is a criterion that can be employed to address that issue. Thus it has gained great popularity as a tool for testing multiple hypotheses. Canonical correlation analysis (CCA) is a statistical technique that is used to make sense of the cross-correlation of two sets of measurements collected on the same set of samples (e.g., brain imaging and genomic data for the same mental illness patients), and sparse CCA extends the classical method to high-dimensional settings. Here, we propose a way of applying the FDR concept to sparse CCA, and a method to control the FDR. The proposed FDR correction directly influences the sparsity of the solution, adapting it to the unknown true sparsity level. Theoretical derivation as well as simulation studies show that our procedure indeed keeps the FDR of the canonical vectors below a user-specified target level. We apply the proposed method to an imaging genomics data set from the Philadelphia Neurodevelopmental Cohort. Our results link the brain connectivity profiles derived from brain activity during an emotion identification task, as measured by functional magnetic resonance imaging, to the corresponding subjects' genomic data.

Brain Functional Network in Chronic Asymptomatic Carotid Artery Stenosis and Occlusion: Changes and Compensation

“…Statistical analyses were performed using SPSS software, version 20 (IBM Corp, Armonk, NY). In the rs-fMRI data network analysis, patients were divided into three groups and bilateral cerebral hemispheres were differentiated; repeated measure analysis of variance (rmANOVA) was used with Bonferroni or FDR correction [ 24 ].…”

Brain Imaging Genomics: Integrated Analysis and Machine Learning

“…An empirical study was performed on an ADNI sample to identify multi-SNP-multi-QT associations between ROI-based imaging QTs and 58 SNPs from AD-related genes. Gossmann et al [175] proposed a FDR-corrected SCCA (FDR-SCCA) procedure to introduce an FDR concept to SCCA and develop a method to control FDR. The existing SCCA methods determine the sparsity parameter using model fit criteria, such as cross validation and permutation.…”

A methylation-based prognostic model predicts survival in patients with colorectal cancer

“…DEGs and DMGs were selected by means of Student's t-test as described above. The false discovery rate (FDR) correction of P value was performed by Benjamini-Hochberg method to reduce the high false positive rate caused by multiple comparisons and the screening criteria were FDR <0.05 (13). DMEGs refer to differential expression in tissues as well as differentially methylated genes.…”