Objective Alzheimer disease (AD) is the most common cause of dementia, a syndrome characterized by cognitive impairment severe enough to interfere with activities of daily life. We aimed to conduct a systematic literature review (SLR) of studies that applied machine learning (ML) methods to clinical data derived from electronic health records in order to model risk for progression of AD dementia. Materials and Methods We searched for articles published between January 1, 2010, and May 31, 2020, in PubMed, Scopus, ScienceDirect, IEEE Explore Digital Library, Association for Computing Machinery Digital Library, and arXiv. We used predefined criteria to select relevant articles and summarized them according to key components of ML analysis such as data characteristics, computational algorithms, and research focus. Results There has been a considerable rise over the past 5 years in the number of research papers using ML-based analysis for AD dementia modeling. We reviewed 64 relevant articles in our SLR. The results suggest that majority of existing research has focused on predicting progression of AD dementia using publicly available datasets containing both neuroimaging and clinical data (neurobehavioral status exam scores, patient demographics, neuroimaging data, and laboratory test values). Discussion Identifying individuals at risk for progression of AD dementia could potentially help to personalize disease management to plan future care. Clinical data consisting of both structured data tables and clinical notes can be effectively used in ML-based approaches to model risk for AD dementia progression. Data sharing and reproducibility of results can enhance the impact, adaptation, and generalizability of this research.

The Epidermal Differentiation Complex (EDC) locus comprises a syntenic and linear cluster of genes whose concomitant expression is a hallmark feature of differentiation in the developing skin epidermis. Many of the EDC proteins are cross-linked together to form the cornified envelope, an essential and discrete unit of the mammalian skin barrier. The mechanism underlying coordinate transcriptional activation of the EDC is unknown. Within the human EDC, we identified an epidermal-specific regulatory enhancer, 923, that responded to the developmental and spatio-temporal cues at the onset of epidermal differentiation in the mouse embryo. Comparative chromosomal conformation capture (3C) assays in proliferating and differentiated primary mouse keratinocytes revealed multiple chromatin interactions that were physiologically sensitive between the 923 enhancer and EDC gene promoters and thus depict the dynamic, chromatin topology of the EDC. We elucidate a mechanistic link between c-Jun/AP-1 and 923, whereby AP-1 and 923-mediated EDC chromatin remodeling is required for functional EDC gene activation. Thus, we identify a critical enhancer/transcription factor axis governing the dynamic regulation of the EDC chromatin architecture and gene expression and provide a framework for future studies towards understanding gene regulation in cutaneous diseases.

The epidermal differentiation complex (EDC) locus consists of a cluster of genes important for the terminal differentiation of the epidermis. While early studies identified the functional importance of individual EDC genes, the recognition of the EDC genes as a cluster with its shared biology, homology, and physical linkage was pivotal to later studies that investigated the transcriptional regulation of the locus. Evolutionary conservation of the EDC and the transcriptional activation during epidermal differentiation suggested a cis-regulatory mechanism via conserved noncoding elements or enhancers. This line of pursuit led to the identification of CNE 923, an epidermal-specific enhancer that was found to mediate chromatin remodeling of the EDC in an AP-1 dependent manner. These genomic studies, as well as the advent of high-throughput sequencing and genome engineering techniques, have paved the way for future investigation into enhancer-mediated regulatory networks in cutaneous biology.

Dozens of variants in the photoreceptor-specific transcription factor (TF) CRX are linked with different human blinding diseases that vary in their severity and age of onset. How different variants in a single TF cause a range of pathological phenotypes is not understood. We deployed massively parallel reporter assays (MPRAs) to measure changes to CRX cis-regulatory function in live mouse retinas carrying knock-ins of two phenotypically distinct human disease-causing Crx variants, one in the DNA binding domain (p.R90W) and the other in the transcriptional effector domain (p.E168d2). We found that the effects of CRX variants on global cis-regulatory activity patterns correspond with the severity of their phenotypes. The variants affect similar sets of enhancers but to different degrees. A subset of silencers were converted to enhancers in retinas lacking a functional CRX effector domain, but were unaffected by p.R90W. Episomal MPRA activities of CRX-bound sequences showed some correspondence with chromatin environments at their original genomic loci, including an enrichment of silencers and depletion of strong enhancers among distal elements whose accessibility increases later in retinal development. Many distal silencers were de-repressed by p.E168d2, but not by p.R90W, suggesting that loss of developmentally timed silencing caused by p.E168d2 may contribute to phenotypic differences between the two variants. Our findings indicate that phenotypically distinct disease variants in different domains of CRX have partially overlapping effects on its cis-regulatory function, leading to mis-regulation of similar sets of enhancers, while having a qualitatively different impact on silencers.

Objectives There is much interest in utilizing clinical data for developing prediction models for Alzheimer’s disease (AD) risk, progression, and outcomes. Existing studies have mostly utilized curated research registries, image analysis, and structured electronic health record (EHR) data. However, much critical information resides in relatively inaccessible unstructured clinical notes within the EHR. Materials and Methods We developed a natural language processing (NLP)-based pipeline to extract AD-related clinical phenotypes, documenting strategies for success and assessing the utility of mining unstructured clinical notes. We evaluated the pipeline against gold-standard manual annotations performed by 2 clinical dementia experts for AD-related clinical phenotypes including medical comorbidities, biomarkers, neurobehavioral test scores, behavioral indicators of cognitive decline, family history, and neuroimaging findings. Results Documentation rates for each phenotype varied in the structured versus unstructured EHR. Interannotator agreement was high (Cohen’s kappa = 0.72–1) and positively correlated with the NLP-based phenotype extraction pipeline’s performance (average F1-score = 0.65–0.99) for each phenotype. Discussion We developed an automated NLP-based pipeline to extract informative phenotypes that may improve the performance of eventual machine learning predictive models for AD. In the process, we examined documentation practices for each phenotype relevant to the care of AD patients and identified factors for success. Conclusion Success of our NLP-based phenotype extraction pipeline depended on domain-specific knowledge and focus on a specific clinical domain instead of maximizing generalizability.