MOTIVATION: There are many diseases with established genetic factors, such as Huntington's disease (HD), that are characterized by variable rates of progression. However, beyond the contribution of the known genetic factors - in this case the Huntingtin (HTT) gene - the impact of the full human genome on the natural progression of such diseases throughout a patient's life remains largely unknown. The increased availability of genome wide association (GWA) data in HD gene expansion carriers (HDGECs), combined with the clinical assessment scores on the same set of patients, has provided a perfect opportunity to assess the potentially broader genetic impact on the natural progression of HD. RESULTS: We present a genetics-driven, probabilistic disease progression model designed to identify and investigate the ways in which a range of genetic factors affect the natural progression of HD. When applied to a clinico-genomic HD dataset, our model identified several single nucleotide polymorphisms (SNPs) with previously unreported effects on disease progression that act at distinct stages and with varying magnitudes. This discovery may shed light on the potential mechanistic impact of previously unidentified genes on HD that may have implications for clinical management. As increasing amounts of GWA data become available more generally, we anticipate that this modeling framework will be broadly applicable to other diseases with strong genetic components. AVAILABILITY AND IMPLEMENTATION: The source code for IHDPM is available at https://github.com/BiomedSciAI/IHDPM.

Clinical researchers use disease progression models to understand patient status and characterize progression patterns from longitudinal health records. One approach for disease progression modeling is to describe patient status using a small number of states that represent distinctive distributions over a set of observed measures. Hidden Markov models (HMMs) and its variants are a class of models that both discover these states and make inferences of health states for patients. Despite the advantages of using the algorithms for discovering interesting patterns, it still remains challenging for medical experts to interpret model outputs, understand complex modeling parameters, and clinically make sense of the patterns. To tackle these problems, we conducted a design study with clinical scientists, statisticians, and visualization experts, with the goal to investigate disease progression pathways of chronic diseases, namely type 1 diabetes (T1D), Huntington's disease, Parkinson's disease, and chronic obstructive pulmonary disease (COPD). As a result, we introduce DPVis which seamlessly integrates model parameters and outcomes of HMMs into interpretable and interactive visualizations. In this article, we demonstrate that DPVis is successful in evaluating disease progression models, visually summarizing disease states, interactively exploring disease progression patterns, and building, analyzing, and comparing clinically relevant patient subgroups.

Diabetes mellitus, commonly known as diabetes, is a chronic disease that often results in multiple complications. Risk prediction of diabetes complications is critical for healthcare professionals to design personalized treatment plans for patients in diabetes care for improved outcomes. In this paper, focusing on Type 2 diabetes mellitus (T2DM), we study the risk of developing complications after the initial T2DM diagnosis from longitudinal patient records. We propose a novel multi-task learning approach to simultaneously model multiple complications where each task corresponds to the risk modeling of one complication. Specifically, the proposed method strategically captures the relationships (1) between the risks of multiple T2DM complications, (2) between different risk factors, and (3) between the risk factor selection patterns, which assumes similar complications have similar contributing risk factors. The method uses coefficient shrinkage to identify an informative subset of risk factors from high-dimensional data, and uses a hierarchical Bayesian framework to allow domain knowledge to be incorporated as priors. The proposed method is favorable for healthcare applications because in addition to improved prediction performance, relationships among the different risks and among risk factors are also identified. Extensive experimental results on a large electronic medical claims database show that the proposed method outperforms state-of-the-art models by a significant margin. Furthermore, we show that the risk associations learned and the risk factors identified lead to meaningful clinical insights.

Type 2 diabetes mellitus (T2DM) is a chronic disease that usually results in multiple complications. Early identification of individuals at risk for complications after being diagnosed with T2DM is of significant clinical value. In this paper, we present a new data-driven predictive approach to predict when a patient will develop complications after the initial T2DM diagnosis. We propose a novel survival analysis method to model the time-to-event of T2DM complications designed to simultaneously achieve two important metrics: 1) accurate prediction of event times, and 2) good ranking of the relative risks of two patients. Moreover, to better capture the correlations of time-to-events of the multiple complications, we further develop a multi-task version of the survival model. To assess the performance of these approaches, we perform extensive experiments on patient level data extracted from a large electronic health record claims database. The results show that our new proposed survival analysis approach consistently outperforms traditional survival models and demonstrate the effectiveness of the multi-task framework over modeling each complication independently.

The rapid growth of Electronic Health Records (EHRs), as well as the accompanied opportunities in Data-Driven Healthcare (DDH), has been attracting widespread interests and attentions. Recent progress in the design and applications of deep learning methods has shown promising results and is forcing massive changes in healthcare academia and industry, but most of these methods rely on massive labeled data. In this work, we propose a general deep learning framework which is able to boost risk prediction performance with limited EHR data. Our model takes a modified generative adversarial network namely ehrGAN, which can provide plausible labeled EHR data by mimicking real patient records, to augment the training dataset in a semi-supervised learning manner. We use this generative model together with a convolutional neural network (CNN) based prediction model to improve the onset prediction performance. Experiments on two real healthcare datasets demonstrate that our proposed framework produces realistic data samples and achieves significant improvements on classification tasks with the generated data over several stat-of-the-art baselines.

MOTIVATION: Modern high-throughput biotechnologies such as microarray are capable of producing a massive amount of information for each sample. However, in a typical high-throughput experiment, only limited number of samples were assayed, thus the classical 'large p, small n' problem. On the other hand, rapid propagation of these high-throughput technologies has resulted in a substantial collection of data, often carried out on the same platform and using the same protocol. It is highly desirable to utilize the existing data when performing analysis and inference on a new dataset. RESULTS: Utilizing existing data can be carried out in a straightforward fashion under the Bayesian framework in which the repository of historical data can be exploited to build informative priors and used in new data analysis. In this work, using microarray data, we investigate the feasibility and effectiveness of deriving informative priors from historical data and using them in the problem of detecting differentially expressed genes. Through simulation and real data analysis, we show that the proposed strategy significantly outperforms existing methods including the popular and state-of-the-art Bayesian hierarchical model-based approaches. Our work illustrates the feasibility and benefits of exploiting the increasingly available genomics big data in statistical inference and presents a promising practical strategy for dealing with the 'large p, small n' problem. AVAILABILITY AND IMPLEMENTATION: Our method is implemented in R package IPBT, which is freely available from https://github.com/benliemory/IPBT CONTACT: [email protected]; [email protected] SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Comprehensive risk assessment lies in the core of enabling proactive healthcare delivery systems. In recent years, data-driven predictive modeling approaches have been increasingly recognized as promising techniques to help enhance healthcare quality and reduce cost. In this paper, we propose a data-driven comprehensive risk prediction method, named LINKAGE, which can be used to jointly assess a set of associated risks in support of holistic care management. Our method can not only perform prediction but also discover the relationships among those risks. The advantages of the proposed model include: 1) It can leverage the relationship between risks and domains and achieve better risk prediction performance; 2) It provides a data-driven approach to understand relationship between risks; 3) It leverages the information between risk prediction and risk association learning to regulate the improvement on both parts; 4) It provides flexibility to incorporate domain knowledge in learning risk associations. We validate the effectiveness of the proposed model on synthetic data and a real-world healthcare survey data set.

MOTIVATION: Further advancement of RNA-Seq technology and its application call for the development of effective normalization methods for RNA-Seq data. Currently, different normalization methods are compared and validated by their correlations with a certain gold standard. Gene expression measurements generated by a different technology or platform such as Real-time reverse transcription polymerase chain reaction (qRT-PCR) or Microarray are usually used as the gold standard. Although the current approach is intuitive and easy to implement, it becomes statistically inadequate when the gold standard is also subject to measurement error (ME). Furthermore, the current approach is not informative, because the correlation of a normalization method with a certain gold standard does not provide much information about the exact quality of the normalized RNA-Seq measurements. RESULTS: We propose to use the system of ME models based on qRT-PCR, Microarray and RNA-Seq gene expression data to compare and validate RNA-Seq normalization methods. This approach does not assume the existence of a gold standard. The performance of a normalization method can be characterized by a group of parameters of the system, which are referred to as the performance parameters, and these performance parameters can be consistently estimated. Different normalization methods can thus be compared by comparing their corresponding estimated performance parameters. We applied the proposed approach to compare five existing RNA-Seq normalization methods using the gene expression data of two RNA samples from the microArray Quality Control and Sequencing Quality Control projects and gained much insight about the pros and cons of these methods.

Our objective is to train $p$-norm Multiple Kernel Learning (MKL) and, more generally, linear MKL regularised by the Bregman divergence, using the Sequential Minimal Optimization (SMO) algorithm. The SMO algorithm is simple, easy to implement and adapt, and efficiently scales to large problems. As a result, it has gained widespread acceptance and SVMs are routinely trained using SMO in diverse real world applications. Training using SMO has been a long standing goal in MKL for the very same reasons. Unfortunately, the standard MKL dual is not differentiable, and therefore can not be optimised using SMO style co-ordinate ascent. In this paper, we demonstrate that linear MKL regularised with the $p$-norm squared, or with certain Bregman divergences, can indeed be trained using SMO. The resulting algorithm retains both simplicity and efficiency and is significantly faster than the state-of-the-art specialised $p$-norm MKL solvers. We show that we can train on a hundred thousand kernels in approximately seven minutes and on fifty thousand points in less than half an hour on a single core.