Discovery of optimal cell type classification marker genes from single cell RNA sequencing data.

The use of single cell/nucleus RNA sequencing (scRNA-seq) technologies that quantitively describe cell transcriptional phenotypes is revolutionizing our understanding of cell biology, leading to new insights in cell type identification, disease mechanisms, and drug development. The tremendous growth in scRNA-seq data has posed new challenges in efficiently characterizing data-driven cell types and identifying quantifiable marker genes for cell type classification. The use of machine learning and explainable artificial intelligence has emerged as an effective approach to study large-scale scRNA-seq data. NS-Forest is a random forest machine learning-based algorithm that aims to provide a scalable data-driven solution to identify minimum combinations of necessary and sufficient marker genes that capture cell type identity with maximum classification accuracy. Here, we describe the latest version, NS-Forest version 4.0 and its companion Python package ( https://github.com/JCVenterInstitute/NSForest ), with several enhancements to select marker gene combinations that exhibit highly selective expression patterns among closely related cell types and more efficiently perform marker gene selection for large-scale scRNA-seq data atlases with millions of cells. By modularizing the final decision tree step, NS-Forest v4.0 can be used to compare the performance of user-defined marker genes with the NS-Forest computationally-derived marker genes based on the decision tree classifiers. To quantify how well the identified markers exhibit the desired pattern of being exclusively expressed at high levels within their target cell types, we introduce the On-Target Fraction metric that ranges from 0 to 1, with a metric of 1 assigned to markers that are only expressed within their target cell types and not in cells of any other cell types. NS-Forest v4.0 outperforms previous versions on its ability to identify markers with higher On-Target Fraction values for closely related cell types and outperforms other marker gene selection approaches at classification with significantly higher F-beta scores when applied to datasets from three human organs - brain, kidney, and lung.

Development of a Medical Complexity Score for Pediatric Aerodigestive Patients.

OBJECTIVE: To develop a complexity scoring system to characterize the diverse population served in pediatric aerodigestive clinics and help predict their treatment outcomes. STUDY DESIGN: A 7-point medical complexity score was developed through an iterative group consensus of relative stakeholders to capture the spectrum of comorbidities among the aerodigestive population. One point was assigned for each comorbid diagnosis in the following categories: airway anomaly, neurologic, cardiac, respiratory, gastrointestinal, genetic diagnoses, and prematurity. A retrospective chart review was conducted of patients seen in the aerodigestive clinic who had ≥2 visits between 2017 and 2021. The predictive value of the complexity score for the selected outcome of feeding progression among children with dysphagia was analyzed with univariate and multivariable logistic regression. RESULTS: We analyzed 234 patients with complexity scores assigned, showing a normal distribution (Shapiro Wilk P = .406) of the scores 1-7 (median, 4; mean, 3.50 ± 1.47). In children with dysphagia, there was waning success in the improvement of oral feeding with increasing complexity scores (OR, 0.66; 95% CI, 0.51-0.84; P = .001). Tube-fed children with higher complexity scores were incrementally less likely to achieve full oral diet (OR, 0.60; 95% CI, 0.40-0.89; P = .01). On multivariable analysis, neurologic comorbidity (OR, 0.26; P < .001) and airway malformation (OR, 0.35; P = .01) were associated with a decreased likelihood to improve in oral feeding. CONCLUSIONS: We propose a novel complexity score for the pediatric aerodigestive population that is easy to use, successfully stratifies diverse presentations, and shows promise as a predictive tool to assist in counseling and resource use.

Leveraging Electronic Health Records for Guideline-Based Asthma Documentation.

BACKGROUND: Asthma is the most common pediatric chronic disease; thus, clinical guidelines have been developed for its assessment and management, which rely on systematic symptom documentation. Electronic health records (EHR) have the potential to record clinical data systematically; however, variability in documentation persists. OBJECTIVE: To identify if the use of a structured asthma template is associated with increased guideline-based asthma documentation and clinical outcomes when compared with the use of nonstructured ones. METHODS: We performed a retrospective case-control study comparing the use of nonstructured templates (NSTs) and asthma-structured templates (ASTs) in new patient and first follow-up encounters, evaluated by pediatric pulmonologists between March 2016 and December 2021. Asthma history items were selected following clinical guidelines, summarized in 29 items for new and 22 items for follow-up encounters. Associations with demographic, spirometry, and health care utilization were explored. RESULTS: A total of 546 initial encounters were included; 450 used structured templates. The use of an AST was associated with higher documentation of asthma items in initial and follow-up encounters. Linear regression analysis showed that the use of ASTs was associated with a 28.2% and 39.65% increase in asthma history completeness (in initial and follow-up encounters, respectively), compared with the use of NSTs. AST use was associated with higher rates of systemic steroid prescriptions within 12 months. No other differences were observed after adjusting for asthma severity. CONCLUSIONS: Using asthma-specific structured templates was associated with increased guideline-based asthma documentation. Leveraging the EHR as a clinical and research tool has the potential to improve clinical practice.

A single-cell regulatory map of postnatal lung alveologenesis in humans and mice.

Ex-utero regulation of the lungs' responses to breathing air and continued alveolar development shape adult respiratory health. Applying single-cell transposome hypersensitive site sequencing (scTHS-seq) to over 80,000 cells, we assembled the first regulatory atlas of postnatal human and mouse lung alveolar development. We defined regulatory modules and elucidated new mechanistic insights directing alveolar septation, including alveolar type 1 and myofibroblast cell signaling and differentiation, and a unique human matrix fibroblast population. Incorporating GWAS, we mapped lung function causal variants to myofibroblasts and identified a pathogenic regulatory unit linked to lineage marker FGF18, demonstrating the utility of chromatin accessibility data to uncover disease mechanism targets. Our regulatory map and analysis model provide valuable new resources to investigate age-dependent and species-specific control of critical developmental processes. Furthermore, these resources complement existing atlas efforts to advance our understanding of lung health and disease across the human lifespan.

Epigenetic Regulation of Myofibroblast Phenotypes in Fibrosis.

PURPOSE OF REVIEW: Myofibroblasts are the fundamental drivers of fibrosing disorders; there is great value in better defining epigenetic networks involved in myofibroblast behavior. Complex epigenetic paradigms, which are likely organ and/or disease specific, direct pathologic myofibroblast phenotypes. In this review, we highlight epigenetic regulators and the mechanisms through which they shape myofibroblast phenotype in fibrotic diseases of different organs. RECENT FINDINGS: Hundreds of genes and their expression contribute to the myofibroblast transcriptional regime influencing myofibroblast phenotype. An increasingly large number of epigenetic modifications have been identified in the regulation of these signaling pathways driving myofibroblast activation and disease progression. Drugs that inhibit or reverse profibrotic epigenetic modifications have shown promise in vitro and in vivo; however, no current epigenetic therapies have been approved to treat fibrosis. Newly described epigenetic mechanisms will be mentioned, along with potential therapeutic targets and innovative strategies to further understand myofibroblast-directed fibrosis. SUMMARY: Epigenetic regulators that direct myofibroblast behavior and differentiation into pathologic myofibroblast phenotypes in fibrotic disorders comprise both overlapping and organ-specific epigenetic mechanisms.