Significant out-of-sample classification from methylation profile scoring for amyotrophic lateral sclerosis

Abstract

We conducted DNA methylation association analyses using Illumina 450K data from whole blood for an Australian amyotrophic lateral sclerosis (ALS) case-control cohort (782 cases and 613 controls). Analyses used mixed linear models as implemented in the OSCA software. We found a significantly higher proportion of neutrophils in cases compared to controls which replicated in an independent cohort from the Netherlands (1159 cases and 637 controls). The OSCA MOMENT linear mixed model has been shown in simulations to best account for confounders. When combined in a methylation profile score, the 25 most-associated probes identified by MOMENT significantly classified case-control status in the Netherlands sample (area under the curve, AUC = 0.65, CI 95% = [0.62-0.68], p = 8.3 × 10 −22). The maximum AUC achieved was 0.69 (CI 95% = [0.66-0.71], p = 4.3 × 10 −34) when cell-type proportion was included in the predictor.

Meta-analysis of genome-wide DNA methylation identifies shared associations across neurodegenerative disorders

“…To discover differentially methylated positions (DMPs) between cases and controls, we conducted MOMENT and MOA MWAS for each available cohort (Additional file 1 : Figures S2, and S3, respectively). We have previously applied both MOA and MOMENT to two independently collected ALS case-control cohorts [ 20 ]. Both methods showed higher out-of-sample classification accuracy compared to linear regression, with MOMENT showing the best performance, despite detection of fewer significantly associated probes.…”

“…Australian ALS cohort The Australian ALS cohort (AUS) and DNA methylation assays have been previously described elsewhere [20]. Part of the Australian sample comprised patients and controls that were ascertained from the University of Sydney as part of the Australian MND DNA bank, which recruited participants from April 2000 to June 2011.…”

“…Participants with AD (N = 211) had neuropsychological profiles which were consistent with AD and were more impaired than participants with mild cognitive impairment (MCI) (N = 133) or healthy controls (N = 768), who performed within expected norms for age on neuropsychological testing. A subset of the AIBL cohort (N = 162 AD cases, N = 94 MCI cases, and N = 485 controls) were subjected to a similar DNA methylation assay protocol as the AUS ALS cohort [20], but bisulfite DNA samples were hybridized to the 8 sample, HumanMethylationEPIC BeadChip Array. The Alzheimer's Disease Neuroimaging Initiative (ADNI) ADNI is a consortium of universities and medical centers in the United States and Canada.…”

“…DNA methylation can repress transcription directly, by inhibiting the binding of specific transcription factors, or indirectly, by recruiting methyl-CpG-binding proteins and their repressive chromatin remodeling activities. There has been increasing evidence that alterations in DNA methylation play an important role in neurodegenerative and other brain disorders, with reports of significant associations with Alzheimer's disease (AD) [13][14][15][16], Parkinson's disease (PD) [17,18], amyotrophic lateral sclerosis (ALS) [19,20], and schizophrenia [21,22], both in the brain and blood. Additionally, DNA methylation data can capture signatures of unmeasured environmental exposures.…”

“…Both methods have been shown, through extensive simulations [ 24 ], to better account for known (cell type proportion, smoking, age, batch effects) and unknown confounders than other methods. We have previously applied both MOA and MOMENT to two independently collected ALS case-control cohorts [ 20 ]. Both methods showed higher out-of-sample classification accuracy compared to linear regression, with MOMENT showing the best performance, despite detection of fewer significantly associated probes.…”