Among the more than 90 identified genetic risk loci for late-onset Alzheimer's disease (AD) and related dementias, the apolipoprotein E (APOE) gene ɛ2/ɛ3/ɛ4 polymorphisms remain the longstanding benchmark for genetic disease risk with a consistently large effect across studies1-10. Despite this massive signal, the exact mechanisms by which ɛ4 increases and ɛ2 decreases dementia risk remain poorly understood. Notably, recent trials of anti-amyloid therapies suggest less efficacy and higher risks of severe side effects in ε4 carriers11-13, hampering the treatment of those with the highest unmet need. To improve our understanding of the genetic architecture of AD in the context of its main genetic driver, we performed genome-wide association studies (GWASs) stratified by ε4 and ε2 carrier status. HP1BP3, SLC50A1, PTPRC, NPAS3, DDHD1, CHST9, SMYD2, PRAMEF1 and GFRA1 emerged as new genomic signals for AD risk, appearing only when stratified by APOE carrier status. DDHD1 appeared especially promising, showing protective effects in ε4 carriers, being identified as an expression quantitative trait locus and being involved in rare neuronal diseases. Such APOE-stratified insights may help understand and overcome side effects, inform clinical trial enrollment strategies, and create the scientific basis for targeted, mechanism-driven therapies in neurodegenerative diseases.
J. Thomassen, H. Leonard, Brittany Ulms et al.· Nature Genetics· 0 citations
Abstract Motivation Copy Number Variations (CNVs) play pivotal roles in complex disease etiology, often requiring large sample sizes to analyze disease associations. While genotyping arrays offer a cost-effective approach for CNV detection using Log R Ratio (LRR) and B Allele Frequency (BAF) signals, existing independent array-based callers suffer from high false positive rates and noise susceptibility, burdening manual validation. Results We present CNV-Finder, a deep learning pipeline employing Long Short-Term Memory (LSTM) networks for large-scale CNV identification within user-defined genomic regions. Trained on expert-annotated samples from the Global Parkinson’s Genetics Program across four neurodegenerative disease-associated genes (PRKN, LINGO2, MAPT, SNCA), CNV-Finder integrates human feedback to iteratively improve performance. In benchmarking across 105 936 samples spanning 11 ancestries and nearly 150 cohorts, the model achieved 91% and 89% visual confirmation rates for PRKN deletions and duplications at high-confidence thresholds. In two validation cohorts, CNV-Finder nominated 83% fewer candidates than a popular Hidden Markov Model-based caller while maintaining higher confirmation rates. Validation through MLPA, short-read, and long-read sequencing demonstrated robust performance, generalizing to diverse signatures including homozygous deletions and SNCA triplications absent from training. Our findings highlight human expertise’s value in complex loci like 17q21.31. Availability and implementation CNV-Finder is freely available at https://github.com/nvk23/CNV-Finder.
Nicole Kuznetsov, Kensuke Daida, M. Makarious et al.· Bioinformatics Advances· 0 citations
