Shaohong Zhang

Generalized Adjusted Rand Indices for cluster ensembles

In this paper, Adjusted Rand Index (ARI) is generalized to two new measures based on matrix comparison: (i) Adjusted Rand Index between a similarity matrix and a cluster partition (ARImp), to evaluate the consistency of a set of clustering solutions with their corresponding consensus matrix in a cluster ensemble, and (ii) Adjusted Rand Index between similarity matrices (ARImm), to evaluate the consistency between two similarity matrices. Desirable properties of ARI are preserved in the two new measures, and new properties are discussed. These properties include: (i) detection of uncorrelatedness; (ii) computation of ARImp/ARImm in a distributed environment; and (iii) characterization of the degree of uncertainty of a consensus matrix. All of these properties are investigated from both the perspectives of theoretical analysis and experimental validation. We have also performed a number of experiments to show the usefulness and effectiveness of the two proposed measures in practical applications.

Transcriptome-Guided Functional Analyses Reveal Novel Biological Properties and Regulatory Hierarchy of Human Embryonic Stem Cell-Derived Ventricular Cardiomyocytes Crucial for Maturation

Abstract Human (h) embryonic stem cells (ESC) represent an unlimited source of cardiomyocytes (CMs); however, these differentiated cells are immature. Thus far, gene profiling studies have been performed with non-purified or non-chamber specific CMs. Here we took a combinatorial approach of using systems biology to guide functional discoveries of novel biological properties of purified hESC-derived ventricular (V) CMs. We profiled the transcriptomes of hESCs, hESC-, fetal (hF) and adult (hA) VCMs, and showed that hESC-VCMs displayed a unique transcriptomic signature. Not only did a detailed comparison between hESC-VCMs and hF-VCMs confirm known expression changes in metabolic and contractile genes, it further revealed novel differences in genes associated with reactive oxygen species (ROS) metabolism, migration and cell cycle, as well as potassium and calcium ion transport. Following these guides, we functionally confirmed that hESC-VCMs expressed IKATP with immature properties, and were accordingly vulnerable to hypoxia/reoxygenation-induced apoptosis. For mechanistic insights, our coexpression and promoter analyses uncovered a novel transcriptional hierarchy involving select transcription factors (GATA4, HAND1, NKX2.5, PPARGC1A and TCF8), and genes involved in contraction, calcium homeostasis and metabolism. These data highlight novel expression and functional differences between hESC-VCMs and their fetal counterparts, and offer insights into the underlying cell developmental state. These findings may lead to mechanism-based methods for in vitro driven maturation.

A New Unsupervised Feature Ranking Method for Gene Expression Data Based on Consensus Affinity

Feature selection is widely established as one of the fundamental computational techniques in mining microarray data. Due to the lack of categorized information in practice, unsupervised feature selection is more practically important but correspondingly more difficult. Motivated by the cluster ensemble techniques, which combine multiple clustering solutions into a consensus solution of higher accuracy and stability, recent efforts in unsupervised feature selection proposed to use these consensus solutions as oracles. However, these methods are dependent on both the particular cluster ensemble algorithm used and the knowledge of the true cluster number. These methods will be unsuitable when the true cluster number is not available, which is common in practice. In view of the above problems, a new unsupervised feature ranking method is proposed to evaluate the importance of the features based on consensus affinity. Different from previous works, our method compares the corresponding affinity of each feature between a pair of instances based on the consensus matrix of clustering solutions. As a result, our method alleviates the need to know the true number of clusters and the dependence on particular cluster ensemble approaches as in previous works. Experiments on real gene expression data sets demonstrate significant improvement of the feature ranking results when compared to several state-of-the-art techniques.

Proteomic Analysis of Human Pluripotent Stem Cell-Derived, Fetal, and Adult Ventricular Cardiomyocytes Reveals Pathways Crucial for Cardiac Metabolism and Maturation

Background—Differentiation of pluripotent human embryonic stem cells (hESCs) to the cardiac lineage represents a potentially unlimited source of ventricular cardiomyocytes (VCMs), but hESC-VCMs are developmentally immature. Previous attempts to profile hESC-VCMs primarily relied on transcriptomic approaches, but the global proteome has not been examined. Furthermore, most hESC-CM studies focus on pathways important for cardiac differentiation, rather than regulatory mechanisms for CM maturation. We hypothesized that gene products and pathways crucial for maturation can be identified by comparing the proteomes of hESCs, hESC-derived VCMs, human fetal and human adult ventricular and atrial CMs. Methods and Results—Using two-dimensional–differential-in-gel electrophoresis, 121 differentially expressed (>1.5-fold; P<0.05) proteins were detected. The data set implicated a role of the peroxisome proliferator–activated receptor &agr; signaling in cardiac maturation. Consistently, WY-14643, a peroxisome proliferator–activated receptor &agr; agonist, increased fatty oxidative enzyme level, hyperpolarized mitochondrial membrane potential and induced a more organized morphology. Along this line, treatment with the thyroid hormone triiodothyronine increased the dynamic tension developed in engineered human ventricular cardiac microtissue by 3-fold, signifying their maturation. Conclusions—We conclude that the peroxisome proliferator–activated receptor &agr; and thyroid hormone pathways modulate the metabolism and maturation of hESC-VCMs and their engineered tissue constructs. These results may lead to mechanism-based methods for deriving mature chamber-specific CMs.

A new method for measuring the semantic similarity on gene ontology

Semantic similarity defined on Gene Ontology (GO) aims to provide the functional relationship between different biological processes, molecular functions, or cellular components. In this paper, a novel method, namely the Shortest Path (SP) algorithm, for measuring the semantic similarity on GO is proposed based on both the GO structure information and the terms property. The proposed algorithm searches for the shortest path that connects two terms and uses the sum of weights on the shortest path to compute the semantic similarity for GO terms. A method for evaluating the nonlinear correlation between two variables is also introduced for validation. Extensive experiments conducted on two public gene expression datasets demonstrate the overall superiority of SP method over the other state-of-the-art methods evaluated.

Image Segmentation Based on Cluster Ensemble

Image segmentation is a classical problem in the area of image processing, multimedia, medical image, and so on. Although there exist a lot of approaches to perform image segmentation, few of them study the image segmentation by the cluster ensemble approach. In this paper, we propose a new algorithm called the cluster ensemble algorithm (CEA) for image segmentation. Specifically, CEA first obtains two set of segmented regions which are partitioned by EM according to the color feature and the texture feature respectively. Then, it integrates these regions to ksegmented regions based on the similarity measure and the fuzzy membership function. Finally, CEA performs the denoise algorithm on the segmented regions to remove the noise. The experiments show that CEA works well during the process of image segmentation.

Generalized Pair-Counting Similarity Measures for Clustering and Cluster Ensembles

In this paper, a number of pair-counting similarity measures associated with a general formulation of cluster ensemble are proposed. These measures are formulated based on our motivation to evaluate the consistency between an individual clustering solution and a cluster ensemble solution, or that between different cluster ensemble solutions, in a uniform manner. A number of criteria are proposed for the comparison of these generalized measures, from both the perspectives of theoretical analysis and experimental validation. We identify their different behaviors and their correlations in different scenarios of traditional clustering solutions and cluster ensembles, with the hope that the results of these studies could 1) serve as important criteria for the design and selection of evaluation measures for clustering solutions, and 2) provide explanations for ambiguous clustering results in related scenarios. Experiments with both synthetic and real data sets are conducted to verify our findings.

Semi-supervised clustering with pairwise and size constraints

In recent years, semi-supervised clustering receives considerable attention in the pattern recognition and data mining communities. This type of clustering algorithms takes advantage of partial prior knowledge, and significant improved performance beyond traditional unsupervised clustering algorithms is observed. In general, the partial prior knowledge is mainly in the form of pairwise constraints, which specify whether point pairs should be in the same cluster or in different clusters. Moreover, some other forms of constraints also attract research interests, for example, the balance constraint or the size constraint. However, it is also important to consider different types of constraints simultaneously, since different types of prior knowledge might have their own bias when considered separately. In this paper, we propose an improved algorithm to incorporate the pairwise and size constraints into a unified framework. Experiments on several benchmark data sets demonstrate that the proposed unified algorithm outperforms previous approaches under a variety of different conditions, which demonstrates that judicious integration of different types of constraints can result in improved performance than in those cases where only a single kind of constraint is used.

Active constrained clustering with multiple cluster representatives

Constrained clustering has recently become an active research topic. This type of clustering methods takes advantage of partial knowledge in the form of pairwise constraints, and acquires significant improvement beyond the traditional un-supervised clustering. However, most of the existing constrained clustering methods use constraints which are selected at random. Recently active constrained clustering algorithms utilizing active constraints have proved themselves to be more effective and efficient. In this paper, we propose an improved algorithm which introduces multiple representatives into constrained clustering to make further use of the active constraints. Experiments on several benchmark data sets and public image data sets demonstrate the advantages of our algorithm over the referenced competitors.

Characterisation of semantic similarity on gene ontology based on a shortest path approach

Semantic similarity defined on Gene Ontology (GO) aims to provide the functional relationship between different GO terms. In this paper, a novel method, namely the Shortest Path (SP) algorithm, for measuring the semantic similarity on GO terms is proposed based on both GO structure information and the terms property. The proposed algorithm searches for the shortest path that connects two terms and uses the sum of weights on the path to estimate the semantic similarity between GO terms. A method for evaluating the nonlinear correlation between two variables is also introduced for validation. Extensive experiments conducted on the PPI dataset and two public gene expression datasets demonstrate the overall superiority of SP method over the other state-of-the-art methods evaluated.