Biomarker Selection for Adaptive Systems.

Biomarkers enable objective monitoring of a given cell or state in a biological system and are widely used in research, biomanufacturing, and clinical practice. However, identifying appropriate biomarkers that are both robustly measurable and capture a state accurately remains challenging. We present a framework for biomarker identification based upon observability guided sensor selection. Our methods, Dynamic Sensor Selection (DSS) and Structure-Guided Sensor Selection (SGSS), utilize temporal models and experimental data, offering a template for applying observability theory to unconventional data obtained from biological systems. Unlike conventional methods that assume well-known, fixed dynamics, DSS adaptively select biomarkers or sensors that maximize observability while accounting for the time-varying nature of biological systems. Additionally, SGSS incorporates structural information and diverse data to identify sensors which are resilient against inaccuracies in our model of the underlying system. We validate our approaches by performing estimation on high dimensional systems derived from temporal gene expression data from partial observations.

Machine learning-based donor permission extraction from informed consent documents.

BACKGROUND: With more clinical trials are offering optional participation in the collection of bio-specimens for biobanking comes the increasing complexity of requirements of informed consent forms. The aim of this study is to develop an automatic natural language processing (NLP) tool to annotate informed consent documents to promote biorepository data regulation, sharing, and decision support. We collected informed consent documents from several publicly available sources, then manually annotated them, covering sentences containing permission information about the sharing of either bio-specimens or donor data, or conducting genetic research or future research using bio-specimens or donor data. RESULTS: We evaluated a variety of machine learning algorithms including random forest (RF) and support vector machine (SVM) for the automatic identification of these sentences. 120 informed consent documents containing 29,204 sentences were annotated, of which 1250 sentences (4.28%) provide answers to a permission question. A support vector machine (SVM) model achieved a F-1 score of 0.95 on classifying the sentences when using a gold standard, which is a prefiltered corpus containing all relevant sentences. CONCLUSIONS: This study provides the feasibility of using machine learning tools to classify permission-related sentences in informed consent documents.

A lipid-associated macrophage lineage rewires the spatial landscape of adipose tissue in early obesity.

Adipose tissue macrophage (ATM) infiltration is associated with adipose tissue dysfunction and insulin resistance in mice and humans. Recent single-cell data highlight increased ATM heterogeneity in obesity but do not provide a spatial context for ATM phenotype dynamics. We integrated single-cell RNA-Seq, spatial transcriptomics, and imaging of murine adipose tissue in a time course study of diet-induced obesity. Overall, proinflammatory immune cells were predominant in early obesity, whereas nonresident antiinflammatory ATMs predominated in chronic obesity. A subset of these antiinflammatory ATMs were transcriptomically intermediate between monocytes and mature lipid-associated macrophages (LAMs) and were consistent with a LAM precursor (pre-LAM). Pre-LAMs were spatially associated with early obesity crown-like structures (CLSs), which indicate adipose tissue dysfunction. Spatial data showed colocalization of ligand-receptor transcripts related to lipid signaling among monocytes, pre-LAMs, and LAMs, including Apoe, Lrp1, Lpl, and App. Pre-LAM expression of these ligands in early obesity suggested signaling to LAMs in the CLS microenvironment. Our results refine understanding of ATM diversity and provide insight into the dynamics of the LAM lineage during development of metabolic disease.

HAT: Hypergraph analysis toolbox.

Recent advances in biological technologies, such as multi-way chromosome conformation capture (3C), require development of methods for analysis of multi-way interactions. Hypergraphs are mathematically tractable objects that can be utilized to precisely represent and analyze multi-way interactions. Here we present the Hypergraph Analysis Toolbox (HAT), a software package for visualization and analysis of multi-way interactions in complex systems.

Evaluating and Extending the Informed Consent Ontology for Representing Permissions from the Clinical Domain.

The purpose of this study was to evaluate, revise, and extend the Informed Consent Ontology (ICO) for expressing clinical permissions, including reuse of residual clinical biospecimens and health data. This study followed a formative evaluation design and used a bottom-up modeling approach. Data were collected from the literature on US federal regulations and a study of clinical consent forms. Eleven federal regulations and fifteen permission-sentences from clinical consent forms were iteratively modeled to identify entities and their relationships, followed by community reflection and negotiation based on a series of predetermined evaluation questions. ICO included fifty-two classes and twelve object properties necessary when modeling, demonstrating appropriateness of extending ICO for the clinical domain. Twenty-six additional classes were imported into ICO from other ontologies, and twelve new classes were recommended for development. This work addresses a critical gap in formally representing permissions clinical permissions, including reuse of residual clinical biospecimens and health data. It makes missing content available to the OBO Foundry, enabling use alongside other widely-adopted biomedical ontologies. ICO serves as a machine-interpretable and interoperable tool for responsible reuse of residual clinical biospecimens and health data at scale.

Deciphering multi-way interactions in the human genome.

Chromatin architecture, a key regulator of gene expression, can be inferred using chromatin contact data from chromosome conformation capture, or Hi-C. However, classical Hi-C does not preserve multi-way contacts. Here we use long sequencing reads to map genome-wide multi-way contacts and investigate higher order chromatin organization in the human genome. We use hypergraph theory for data representation and analysis, and quantify higher order structures in neonatal fibroblasts, biopsied adult fibroblasts, and B lymphocytes. By integrating multi-way contacts with chromatin accessibility, gene expression, and transcription factor binding, we introduce a data-driven method to identify cell type-specific transcription clusters. We provide transcription factor-mediated functional building blocks for cell identity that serve as a global signature for cell types.

Functional organization of the maternal and paternal human 4D Nucleome.

Every human somatic cell inherits a maternal and a paternal genome, which work together to give rise to cellular phenotypes. However, the allele-specific relationship between gene expression and genome structure through the cell cycle is largely unknown. By integrating haplotype-resolved genome-wide chromosome conformation capture, mature and nascent mRNA, and protein binding data from a B lymphoblastoid cell line, we investigate this relationship both globally and locally. We introduce the maternal and paternal 4D Nucleome, enabling detailed analysis of the mechanisms and dynamics of genome structure and gene function for diploid organisms. Our analyses find significant coordination between allelic expression biases and local genome conformation, and notably absent expression bias in universally essential cell cycle and glycolysis genes. We propose a model in which coordinated biallelic expression reflects prioritized preservation of essential gene sets.

Understanding memory B cell selection.

The mammalian adaptive immune system has evolved over millions of years to become an incredibly effective defense against foreign antigens. The adaptive immune system's humoral response creates plasma B cells and memory B cells, each with their own immunological objectives. The affinity maturation process is widely viewed as a heuristic to solve the global optimization problem of finding B cells with high affinity to the antigen. However, memory B cells appear to be purposely selected earlier in the affinity maturation process and have lower affinity. We propose that this memory B cell selection process may be an approximate solution to two optimization problems: optimizing for affinity to similar antigens in the future despite mutations or other minor differences, and optimizing to warm start the generation of plasma B cells in the future. We use simulations to provide evidence for our hypotheses, taking into account data showing that certain B cell mutations are more likely than others. Our findings are consistent with memory B cells having high-affinity to mutated antigens, but do not provide strong evidence that memory B cells will be more useful than selected naive B cells for seeding the secondary germinal centers.

Lessons Learned for Identifying and Annotating Permissions in Clinical Consent Forms.

BACKGROUND: The lack of machine-interpretable representations of consent permissions precludes development of tools that act upon permissions across information ecosystems, at scale. OBJECTIVES: To report the process, results, and lessons learned while annotating permissions in clinical consent forms. METHODS: We conducted a retrospective analysis of clinical consent forms. We developed an annotation scheme following the MAMA (Model-Annotate-Model-Annotate) cycle and evaluated interannotator agreement (IAA) using observed agreement (A o), weighted kappa (κw ), and Krippendorff's α. RESULTS: The final dataset included 6,399 sentences from 134 clinical consent forms. Complete agreement was achieved for 5,871 sentences, including 211 positively identified and 5,660 negatively identified as permission-sentences across all three annotators (A o = 0.944, Krippendorff's α = 0.599). These values reflect moderate to substantial IAA. Although permission-sentences contain a set of common words and structure, disagreements between annotators are largely explained by lexical variability and ambiguity in sentence meaning. CONCLUSION: Our findings point to the complexity of identifying permission-sentences within the clinical consent forms. We present our results in light of lessons learned, which may serve as a launching point for developing tools for automated permission extraction.

Expressing and Executing Informed Consent Permissions Using SWRL: The All of Us Use Case.

The informed consent process is a complicated procedure involving permissions as well a variety of entities and actions. In this paper, we discuss the use of Semantic Web Rule Language (SWRL) to further extend the Informed Consent Ontology (ICO) to allow for semantic machine-based reasoning to manage and generate important permission-based information that can later be viewed by stakeholders. We present four use cases of permissions from the All of Us informed consent document and translate these permissions into SWRL expressions to extend and operationalize ICO. Our efforts show how SWRL is able to infer some of the implicit information based on the defined rules, and demonstrate the utility of ICO through the use of SWRL extensions. Future work will include developing formal and generalized rules and expressing permissions from the entire document, as well as working towards integrating ICO into software systems to enhance the semantic representation of informed consent for biomedical research.

Predicting 30-day hospital readmissions using artificial neural networks with medical code embedding.

Reducing unplanned readmissions is a major focus of current hospital quality efforts. In order to avoid unfair penalization, administrators and policymakers use prediction models to adjust for the performance of hospitals from healthcare claims data. Regression-based models are a commonly utilized method for such risk-standardization across hospitals; however, these models often suffer in accuracy. In this study we, compare four prediction models for unplanned patient readmission for patients hospitalized with acute myocardial infarction (AMI), congestive health failure (HF), and pneumonia (PNA) within the Nationwide Readmissions Database in 2014. We evaluated hierarchical logistic regression and compared its performance with gradient boosting and two models that utilize artificial neural networks. We show that unsupervised Global Vector for Word Representations embedding representations of administrative claims data combined with artificial neural network classification models improves prediction of 30-day readmission. Our best models increased the AUC for prediction of 30-day readmissions from 0.68 to 0.72 for AMI, 0.60 to 0.64 for HF, and 0.63 to 0.68 for PNA compared to hierarchical logistic regression. Furthermore, risk-standardized hospital readmission rates calculated from our artificial neural network model that employed embeddings led to reclassification of approximately 10% of hospitals across categories of hospital performance. This finding suggests that prediction models that incorporate new methods classify hospitals differently than traditional regression-based approaches and that their role in assessing hospital performance warrants further investigation.