Integrating hypertension phenotype and genotype with hybrid non-negative matrix factorization

Northwestern University resource and education development initiatives to advance collaborative artificial intelligence across the learning health system

Introduction

The rapid development of artificial intelligence (AI) in healthcare has exposed the unmet need for growing a multidisciplinary workforce that can collaborate effectively in the learning health systems. Maximizing the synergy among multiple teams is critical for Collaborative AI in Healthcare.

Methods

We have developed a series of data, tools, and educational resources for cultivating the next generation of multidisciplinary workforce for Collaborative AI in Healthcare. We built bulk-natural language processing pipelines to extract structured information from clinical notes and stored them in common data models. We developed multimodal AI/machine learning (ML) tools and tutorials to enrich the toolbox of the multidisciplinary workforce to analyze multimodal healthcare data. We have created a fertile ground to cross-pollinate clinicians and AI scientists and train the next generation of AI health workforce to collaborate effectively.

Results

Our work has democratized access to unstructured health information, AI/ML tools and resources for healthcare, and collaborative education resources. From 2017 to 2022, this has enabled studies in multiple clinical specialties resulting in 68 peer-reviewed publications. In 2022, our cross-discipline efforts converged and institutionalized into the Center for Collaborative AI in Healthcare.

Conclusions

Our Collaborative AI in Healthcare initiatives has created valuable educational and practical resources. They have enabled more clinicians, scientists, and hospital administrators to successfully apply AI methods in their daily research and practice, develop closer collaborations, and advanced the institution-level learning health system.

Clustering single-cell multimodal omics data with jrSiCKLSNMF

Introduction: The development of multimodal single-cell omics methods has enabled the collection of data across different omics modalities from the same set of single cells. Each omics modality provides unique information about cell type and function, so the ability to integrate data from different modalities can provide deeper insights into cellular functions. Often, single-cell omics data can prove challenging to model because of high dimensionality, sparsity, and technical noise. Methods: We propose a novel multimodal data analysis method called joint graph-regularized Single-Cell Kullback-Leibler Sparse Non-negative Matrix Factorization (jrSiCKLSNMF, pronounced "junior sickles NMF") that extracts latent factors shared across omics modalities within the same set of single cells. Results: We compare our clustering algorithm to several existing methods on four sets of data simulated from third party software. We also apply our algorithm to a real set of cell line data. Discussion: We show overwhelmingly better clustering performance than several existing methods on the simulated data. On a real multimodal omics dataset, we also find our method to produce scientifically accurate clustering results.

Subtyping based on immune cell fractions reveal heterogeneity of cardiac fibrosis in end-stage heart failure

Background

A central issue hindering the development of effective anti-fibrosis drugs for heart failure is the unclear interrelationship between fibrosis and the immune cells. This study aims at providing precise subtyping of heart failure based on immune cell fractions, elaborating their differences in fibrotic mechanisms, and proposing a biomarker panel for evaluating intrinsic features of patients' physiological statuses through subtype classification, thereby promoting the precision medicine for cardiac fibrosis.

Methods

We inferred immune cell type abundance of the ventricular samples by a computational method (CIBERSORTx) based on ventricular tissue samples from 103 patients with heart failure, and applied K-means clustering to divide patients into two subtypes based on their immune cell type abundance. We also designed a novel analytic strategy: Large-Scale Functional Score and Association Analysis (LAFSAA), to study fibrotic mechanisms in the two subtypes.

Results

Two subtypes of immune cell fractions: pro-inflammatory and pro-remodeling subtypes, were identified. LAFSAA identified 11 subtype-specific pro-fibrotic functional gene sets as the basis for personalised targeted treatments. Based on feature selection, a 30-gene biomarker panel (ImmunCard30) established for diagnosing patient subtypes achieved high classification performance, with the area under the receiver operator characteristic curve corresponding to 0.954 and 0.803 for the discovery and validation sets, respectively.

Conclusion

Patients with the two subtypes of cardiac immune cell fractions were likely having different fibrotic mechanisms. Patients' subtypes can be predicted based on the ImmunCard30 biomarker panel. We envision that our unique stratification strategy revealed in this study will unravel advance diagnostic techniques for personalised anti-fibrotic therapy.

Multimodal machine learning in precision health: A scoping review

Machine learning is frequently being leveraged to tackle problems in the health sector including utilization for clinical decision-support. Its use has historically been focused on single modal data. Attempts to improve prediction and mimic the multimodal nature of clinical expert decision-making has been met in the biomedical field of machine learning by fusing disparate data. This review was conducted to summarize the current studies in this field and identify topics ripe for future research. We conducted this review in accordance with the PRISMA extension for Scoping Reviews to characterize multi-modal data fusion in health. Search strings were established and used in databases: PubMed, Google Scholar, and IEEEXplore from 2011 to 2021. A final set of 128 articles were included in the analysis. The most common health areas utilizing multi-modal methods were neurology and oncology. Early fusion was the most common data merging strategy. Notably, there was an improvement in predictive performance when using data fusion. Lacking from the papers were clear clinical deployment strategies, FDA-approval, and analysis of how using multimodal approaches from diverse sub-populations may improve biases and healthcare disparities. These findings provide a summary on multimodal data fusion as applied to health diagnosis/prognosis problems. Few papers compared the outputs of a multimodal approach with a unimodal prediction. However, those that did achieved an average increase of 6.4% in predictive accuracy. Multi-modal machine learning, while more robust in its estimations over unimodal methods, has drawbacks in its scalability and the time-consuming nature of information concatenation.

Multi-Trait Genetic Analysis Reveals Clinically Interpretable Hypertension Subtypes

Background:

Hypertension comprises a heterogeneous range of phenotypes. We asked whether underlying genetic structure could explain a part of this heterogeneity.

Methods:

Our study sample comprised N=198 148 FinnGen participants (56% women, mean age 58 years) and N=21 168 well-phenotyped FINRISK participants (53% women, mean age 50 years). First, we identified genetic hypertension components with an unsupervised Bayesian non-negative matrix factorization algorithm using public genome-wide association data for 144 genetic hypertension variants and 16 clinical traits. For these components, we computed their (1) cross-sectional associations with clinical traits in FINRISK using linear regression and (2) longitudinal associations with incident adverse outcomes in FinnGen using Cox regression.

Results:

We observed 4 genetic hypertension components corresponding to recognizable clinical phenotypes: obesity (high body mass index), dyslipidemia (low high-density lipoprotein cholesterol and high triglycerides), hypolipidemia (low low-density lipoprotein cholesterol and low total cholesterol), and short stature. In FINRISK, all hypertension components had robust associations with their respective clinical characteristics. In FinnGen, the Obesity component was associated with increased diabetes risk (hazard ratio per 1 SD increase 1.08 [Bonferroni corrected CI, 1.05–1.10]) and the Hypolipidemia component with increased autoimmune disease risk (hazard ratio per 1 SD increase 1.05 [Bonferroni corrected CI, 1.03–1.07]). In addition, all hypertension components were related to both hypertension and cardiovascular disease.

Conclusions:

Our unsupervised analysis demonstrates that the genetic basis of hypertension can be understood as a mixture of 4 broad, clinically interpretable components capturing disease heterogeneity. These components could be used to stratify individuals into specific genetic subtypes and, therefore, to benefit personalized health care and pharmaceutical research.

AX-Unet: A Deep Learning Framework for Image Segmentation to Assist Pancreatic Tumor Diagnosis

Image segmentation plays an essential role in medical imaging analysis such as tumor boundary extraction. Recently, deep learning techniques have dramatically improved performance for image segmentation. However, an important factor preventing deep neural networks from going further is the information loss during the information propagation process. In this article, we present AX-Unet, a deep learning framework incorporating a modified atrous spatial pyramid pooling module to learn the location information and to extract multi-level contextual information to reduce information loss during downsampling. We also introduce a special group convolution operation on the feature map at each level to achieve information decoupling between channels. In addition, we propose an explicit boundary-aware loss function to tackle the blurry boundary problem. We evaluate our model on two public Pancreas-CT datasets, NIH Pancreas-CT dataset, and the pancreas part in medical segmentation decathlon (MSD) medical dataset. The experimental results validate that our model can outperform the state-of-the-art methods in pancreas CT image segmentation. By comparing the extracted feature output of our model, we find that the pancreatic region of normal people and patients with pancreatic tumors shows significant differences. This could provide a promising and reliable way to assist physicians for the screening of pancreatic tumors.

Globally learning gene regulatory networks based on hidden atomic regulators from transcriptomic big data

BACKGROUND

Genes are regulated by various types of regulators and most of them are still unknown or unobserved. Current gene regulatory networks (GRNs) reverse engineering methods often neglect the unknown regulators and infer regulatory relationships in a local and sub-optimal manner.

RESULTS

This paper proposes a global GRNs inference framework based on dictionary learning, named dlGRN. The method intends to learn atomic regulators (ARs) from gene expression data using a modified dictionary learning (DL) algorithm, which reflects the whole gene regulatory system, and predicts the regulation between a known regulator and a target gene in a global regression way. The modified DL algorithm fits the scale-free property of biological network, rendering dlGRN intrinsically discern direct and indirect regulations.

CONCLUSIONS

Extensive experimental results on simulation and real-world data demonstrate the effectiveness and efficiency of dlGRN in reverse engineering GRNs. A novel predicted transcription regulation between a TF TFAP2C and an oncogene EGFR was experimentally verified in lung cancer cells. Furthermore, the real application reveals the prevalence of DNA methylation regulation in gene regulatory system. dlGRN can be a standalone tool for GRN inference for its globalization and robustness.

Conceptualization of Heterogeneity of Chronic Diseases and Atherosclerosis as a Pathway to Precision Medicine: Endophenotype, Endotype, and Residual Cardiovascular Risk

The article discusses modern approaches to the conceptualization of pathogenetic heterogeneity in various branches of medical science. The concepts of endophenotype, endotype, and residual cardiovascular risk and the scope of their application in internal medicine and cardiology are considered. Based on the latest results of studies of the genetic architecture of atherosclerosis, five endotypes of atherosclerosis have been proposed. Each of the presented endotypes represents one or another pathophysiological mechanism of atherogenesis, having an established genetic substrate, a characteristic panel of biomarkers, and a number of clinical features. Clinical implications and perspectives for the study of endotypes of atherosclerosis are briefly reviewed.

Getting to the Cores of Autism

The genetic architecture of autism spectrum disorder (ASD) is itself a diverse allelic spectrum that consists of rare de novo or inherited variants in hundreds of genes and common polygenic risk at thousands of loci. ASD susceptibility genes are interconnected at the level of transcriptional and protein networks, and many function as genetic regulators of neurodevelopment or synaptic proteins that regulate neural activity. So that the core underlying neuropathologies can be further elucidated, we emphasize the importance of first defining subtypes of ASD on the basis of the phenotypic signatures of genes in model systems and humans.

Cancer classification and pathway discovery using non-negative matrix factorization

OBJECTIVES

Extracting genetic information from a full range of sequencing data is important for understanding disease. We propose a novel method to effectively explore the landscape of genetic mutations and aggregate them to predict cancer type.

DESIGN

We applied non-smooth non-negative matrix factorization (nsNMF) and support vector machine (SVM) to utilize the full range of sequencing data, aiming to better aggregate genetic mutations and improve their power to predict disease type. More specifically, we introduce a novel classifier to distinguish cancer types using somatic mutations obtained from whole-exome sequencing data. Mutations were identified from multiple cancers and scored using SIFT, PP2, and CADD, and collapsed at the individual gene level. nsNMF was then applied to reduce dimensionality and obtain coefficient and basis matrices. A feature matrix was derived from the obtained matrices to train a classifier for cancer type classification with the SVM model.

RESULTS

We have demonstrated that the classifier was able to distinguish four cancer types with reasonable accuracy. In five-fold cross-validations using mutation counts as features, the average prediction accuracy was 80% (SEM = 0.1%), significantly outperforming baselines and outperforming models using mutation scores as features.

CONCLUSION

Using the factor matrices derived from the nsNMF, we identified multiple genes and pathways that are significantly associated with each cancer type. This study presents a generic and complete pipeline to study the associations between somatic mutations and cancers. The proposed method can be adapted to other studies for disease status classification and pathway discovery.

Recent Advances in Supervised Dimension Reduction: A Survey

Recently, we have witnessed an explosive growth in both the quantity and dimension of data generated, which aggravates the high dimensionality challenge in tasks such as predictive modeling and decision support. Up to now, a large amount of unsupervised dimension reduction methods have been proposed and studied. However, there is no specific review focusing on the supervised dimension reduction problem. Most studies performed classification or regression after unsupervised dimension reduction methods. However, we recognize the following advantages if learning the low-dimensional representation and the classification/regression model simultaneously: high accuracy and effective representation. Considering classification or regression as being the main goal of dimension reduction, the purpose of this paper is to summarize and organize the current developments in the field into three main classes: PCA-based, Non-negative Matrix Factorization (NMF)-based, and manifold-based supervised dimension reduction methods, as well as provide elaborated discussions on their advantages and disadvantages. Moreover, we outline a dozen open problems that can be further explored to advance the development of this topic.