A unified computational model for revealing and predicting subtle subtypes of cancers
 Xianwen Ren^{1},
 Yong Wang^{2, 3}Email author,
 Jiguang Wang^{4} and
 XiangSun Zhang^{2, 3}Email author
DOI: 10.1186/147121051370
© Ren et al; licensee BioMed Central Ltd. 2012
Received: 10 December 2011
Accepted: 4 April 2012
Published: 1 May 2012
Abstract
Background
Gene expression profiling technologies have gradually become a community standard tool for clinical applications. For example, gene expression data has been analyzed to reveal novel disease subtypes (class discovery) and assign particular samples to welldefined classes (class prediction). In the past decade, many effective methods have been proposed for individual applications. However, there is still a pressing need for a unified framework that can reveal the complicated relationships between samples.
Results
We propose a novel convex optimization model to perform class discovery and class prediction in a unified framework. An efficient algorithm is designed and software named OTCC (Optimization Tool for Clustering and Classification) is developed. Comparison in a simulated dataset shows that our method outperforms the existing methods. We then applied OTCC to acute leukemia and breast cancer datasets. The results demonstrate that our method not only can reveal the subtle structures underlying those cancer gene expression data but also can accurately predict the class labels of unknown cancer samples. Therefore, our method holds the promise to identify novel cancer subtypes and improve diagnosis.
Conclusions
We propose a unified computational framework for class discovery and class prediction to facilitate the discovery and prediction of subtle subtypes of cancers. Our method can be generally applied to multiple types of measurements, e.g., gene expression profiling, proteomic measuring, and recent nextgeneration sequencing, since it only requires the similarities among samples as input.
Keywords
Class discovery Class prediction Quadratic programming CancerBackground
Accurate diagnosis is a great challenge for clinical therapies. In particular, the current diagnosis based on only a few genes, proteins or metabolites are very limited when it comes to tackling the intrinsic complexity of many diseases, e.g., cancers. Fortunately with the rapid development of highthroughput technologies, gene expression profiling techniques have been widely applied in clinical research. The big advantage is to simultaneously measure the expressions of thousands of genes [1–4]. To date, two types of strategies have been widely used to analyze gene expression data for clinical purpose: class discovery and class prediction. Class discovery tries to identify new disease subtypes while class prediction tries to assign particular samples to welldefined disease classes [5]. Both tasks have significant potentials to improve cancer diagnosis, prognosis, and therapies but require effective and efficient computational methods to deal with the large amount of data involved.
In the machine learning framework, class discovery is an unsupervised task. Many methods related to clustering have been proposed and applied to identify new disease subtypes. Several wellknown methods, e.g., hierarchical clustering (HC), selforganizing maps (SOM), and nonnegative matrix factorization (NMF) have been successfully used [6–14]. Class prediction is generally supervised. Supervised methods, e.g., support vector machines, Bayes classifiers, k nearest neighbors, etc., have been adopted [15–19]. However, class discovery and class prediction are by nature closely linked to each other. Their separate computational implementation prevents clinicians from comparing the results obtained in unsupervised and supervised settings. Alexandridis et al. developed a mixture model unifying two tasks and obtained promising results [20]. However, the global optimum of their model cannot be guaranteed in theory and is difficult to obtain in practice. In addition, estimating the mixture distributions often involves profiling the gene expressions of many clinical samples, which is time consuming and also very expensive. Therefore, a universal, easily solvable computational framework is highly desirable to help clinicians understand such diseases using fewer samples.
In this paper, we propose a semisupervised solution to formulate class discovery and class prediction into a unified framework. We term it OTCC (Optimization Tool for Clustering and Classification). The underlying principle is to seek an optimal sample labeling scheme to ensure that similar samples can be assigned with similar labels. This assumption is straightforward and can be easily understood by clinicians. OTCC has several prominent features: 1) The global optimal solution is guaranteed because it is based on convex quadratic programming; 2) It implements class discovery and class prediction in one computational framework; 3) It does not require many samples; 4) It can be applied to both small and large datasets due to a customized iterative algorithm. Experiments on acute leukemia and breast cancer datasets suggest the validity and advantages of OTCC in mining the clinical significance of patient gene expression data.
Methods
Overview of the optimization model
For simplicity, we consider two classes to illustrate the optimization model. We note that both class discovery and class prediction for the two classes can be transformed into a sample labeling problem. In this section, the optimization model is formulated to find the best way to assign labels to the samples. The labeling problem for multiclass cases for class discovery and class prediction will be discussed in the next sections.
where N is the total number of samples; s_{ ij } is the similarity score of samples x_{ i } and x_{ j, } which is calculated from the gene expression profiles; and f_{ i } is the unknown variable to be determined and represents the label of sample x_{ i }. A is a set of samples that are known to belong to Class Zero. B is a set of samples that are known to belong to Class One. The objective function in Equation (1) tends to assign similar labels to similar samples (${s}_{\mathit{ij}}>0$). Constraints in Equation (2) ensure that the resultant sample labels are consistent with the known information and that the final labels ${f}_{i}$ are between zero and one.
The objective function (1) can be rewritten in vector form as ${f}^{T}Lf$. Here f is the sample label vector (f_{i,} is the label of Sample i) and L is the Laplacian matrix of the similarity matrix S (s_{ ij }, the similarity score of samples i and j), i.e., L = D − S and D is a diagonal matrix with ${d}_{\mathit{ii}}=\sum _{j=1}^{\mathit{N}}{s}_{\mathit{ij}}$. If s_{ ij } are all nonnegative, L is positive semidefinite. The objective function is convex and the constraints are linear. Thus the model (1–2) is a convex quadratic programming problem and a global optimal solution is guaranteed.
Due to the form of the objective function, our optimization model is tightly related to spectral clustering and semisupervised learning [21–23]. These links form the basis for class discovery and class prediction. Importantly, the constraints imposed in this model provide a few advantages for cutoff setting and outlier identification.
The sample similarity matrix
Usually the gene expression profile for n genes and m samples is mathematically denoted by an $n\times m$ matrix X. Each element x_{ ij } represents the expression level of gene i in sample j. x_{ i } is an mdimensional vector denoting the expression value of gene i. The construction of the sample similarity matrix is important because it is the only input for model (1–2) to fully utilize the gene expression data. Since the calculation of the similarity matrix and the solving of the optimization model are separated, various feature selection/extraction techniques and different measures of similarity can be applied here to incorporate prior information. A simple and straightforward method to construct a similarity matrix of samples based on the gene expression profiles is to calculate the Pearson correlation coefficients of each sample pair which provides a uniform measure between −1 and 1. To get nonnegative s_{ ij }, a linear transformation can be adopted to map [−1, 1] to [0, 1]. Because the Pearson correlation coefficients based on the gene expression profiles are calculated pairwisely between every two samples, it does not consider the similarities among samples globally. To provide a global similarity measure, a secondorder correlation similarity matrix can be constructed by exploiting the deduced sample correlation features (i.e., calculating the Pearson correlation coefficients of the sample correlation vectors). In this study we used secondorder correlation similarity matrices to identify the underlying structures of cancer gene expression data.
Setting for class discovery
Given the similarity matrix S, sets A and B are necessary to implement the class discovery task through Model (1–2). If A and B are not provided, i.e., without the corresponding constraints in Equation (2), the optimization model results in a trivial solution given nonnegative s_{ ij }. The trivial solution indicates that all the samples belong to one class, which is meaningless. To obtain a meaningful solution, A and B should be specified and intersection between A and B is not allowed. Usually for class discovery task, information about A and B is not available since all sample labels are unknown. Here we introduce a weak assumption to set up A and B. We name it here as the most dissimilar assumption. The assumption is that the two least similar samples should belong to different classes. Otherwise all samples should belong to one class. According to this assumption, the minimal s_{ ij } for $i,j\in \left\{1,\cdots ,N\right\}$ is identified, denoted by s_{ ab }. Let Sample x_{ a } be labeled with zero and x_{ b } be labeled with one, or vice versa. If there is more than one minimal value in S, the sample pair with minimal values in S^{n} (the power of similarity matrix S, where n > 1 is a positive integer) is also a candidate to determine set A and B. Model (1–2) is then well constructed and optimal labeling can be uniquely determined by solving the model.
Setting for class prediction
Class prediction tries to assign a set of particular samples to known classes. In this setting, goldstandard data are generally available and some gene expression profiles for samples are labeled with known classes. That is, A and B are available. Model (1–2) can therefore be implemented for class prediction.
A fast algorithm for largescale problems
Model (1–2) can be considered convex quadratic programming if all values of s_{ ij } are positive. It can be solved efficiently by the general solvers such as quadprog in Matlab and the sequential minimal optimization (SMO) algorithm which has been applied successfully to solve the optimization problems in support vector machine applications. Here, a simple customized algorithm is proposed to solve Model (1–2) quickly, even for very largescale problems by fully considering its particular characteristics.
We design the following algorithm to quickly find the solution:

Step 1: Let $t=0$ and ${f}_{a}=0$ for $a\in A$, ${f}_{b}=1$ for $b\in B$ and ${f}_{i}=0$ for $i\in \left\{1,\cdots ,N\right\}/A/B$.

Step 2: Calculate ${f}_{i}^{t+1}=\frac{\sum _{j=1}^{\mathit{N}}{s}_{\mathit{ij}}{f}_{j}^{t}}{\sum _{j=1}^{\mathit{N}}{s}_{\mathit{ij}}}$ for $i\in \left\{1,\cdots ,N\right\}/A/B$.

Step 3: Let $t=t+1$. If $\underset{i}{max}{f}_{i}^{t}{f}_{i}^{t1}$ is less than a predefined threshold or t is larger than the maximal steps allowed, stop; otherwise, repeat Step 2 and Step 3.
Next, we prove the above algorithm is correct and convergent.
Theroem 1: Suppose Algorithm 1 gives rise to the sequence,${f}^{0},{f}^{1},\dots ,{f}^{t},{f}^{t+1},\dots $. It converges to ${f}^{*}$. ${f}^{*}$ satisfies the KKT point of Model (1)(2).
Secondly we show Algorithm 1 is correct. At convergence, the solution is ${f}^{*}$ and satisfies ${f}_{i}^{*}=\frac{\sum _{j=1}^{\mathit{N}}{s}_{\mathit{ij}}{f}_{j}^{*}}{\sum _{j=1}^{\mathit{N}}{s}_{\mathit{ij}}}$ for $i\in \{1,\cdots ,N\}/A/B$. ${f}_{a}^{*}=0$for $a\in A$ and ${f}_{b}^{*}=1$ for $b\in B$also hold. Then ${f}^{*}$ satisfies the KKT condition (4)(5). This proves our algorithm correctly converges to a minimum satisfying KKT condition.
One advantage of our algorithm is that the computational complexity is low and it requires only a small amount of computer memory. So our algorithm can be applied to very large data sets.
Postprocessing the solutions
Each sample gets a continuous label between zero and one after the optimization model (1)(2) is solved. We can easily obtain the binary labels by applying a predefined threshold. If a training data set is available, this threshold can be learned from the training data by crossvalidation. Otherwise, the median of zero and one, 0.5, is a natural cutoff to convert the continuous labels into binary labels. If label ${f}_{i}$ is close to zero, i.e., ${f}_{i}<0.5$, the corresponding sample should be classified to Class Zero. Otherwise, if label ${f}_{i}$ is close to one, i.e., ${f}_{i}>0.5$, the corresponding sample will be classified to Class One. This is a great option compared to traditional spectral clustering methods in which the cutoff needs considerable human intervention. This advantage makes it much easier for clinicians and biologists to use.
Multipleclass cases

Step 1: Classify samples into two classes by OTCC.

Step 2: Calculate the inner minimum similarity score for each class. If the minimum similarity score of some class is less than a predefined threshold, then repeat Step 1 to classify the samples of this class into two subclasses.

Step 3: repeat Step 2 until all the inner minimum similarity scores of the classes are above the threshold.
The procedure does not require the number of clusters but instead relies on the least tolerant similarity score within classes. Compared to the number of clusters which is generally required by many existing class discovery methods, our similarity score is tightly related to the expert’s knowledge and is expected to be defined by clinicians and biologists based on their knowledge. Alternatively, without predefining a stopping criterion, OTCC can be applied recursively until each sample is a single class. This outputs a binary tree in which all samples are leaves and the relationships among them are fully depicted. This property allows OTCC to reveal the fine structure of patient samples.
For class prediction cases, the relationship between multiple classes can be organized as a binary tree and then the model can be applied recursively according to the binary tree to obtain the labels of all samples. The binary tree should reflect the relationship of the classes. Otherwise wrong prior information will be introduced and mislead the class prediction results. When the class relationships are not available or all the classes are independent of each other, an arbitrary binary tree can be used. Onevsone or onevsall strategies can also be adopted to extend OTCC to multiclass cases.
Results and discussion
Performance of OTCC on simulated data sets
The second simulation dataset consisted of multiple classes and was generated using a similar procedure. For multiple classes, we applied OTCC recursively to construct a binary tree to reveal the multiple classes. If the real relationship among multiple classes is indeed a binary tree, it is reasonable to expect OTCC to succeed. Here we consider an extreme example to show that OTCC can also successfully deal with cases in which the relationship among multiple classes is inherently not a binary tree.
The success of OTCC for resolving the above multicluster structure lies in its ability to form pseudoclusters when clustering. There are two globally optimum solutions in this case (Nodes 11 to 15 have the same labels as Nodes 1 to 5 or Nodes 6 to 10). OTCC assigns Nodes 11 to 15 to the same labels as Nodes 1 to 5, generating a degenerative pseudocluster whereas Nodes 6 to 10 are classified correctly first. We recursively applying OTCC to pseudoclusters until the consistence criterion applies to each cluster. In this way it resolves the multicluster structure irrespective of whether the relationship among the multiple classes is inherently a binary tree or not.
In Figure 2A, the three clusters are balanced (with the same number of nodes). We also simulate the unbalanced and noisy data set by changing the number of nodes within clusters and adding betweencluster links. OTCC can still resolve the underlying multicluster structure (Figure 2B).
Experiments on cancer gene expression data sets
Next we use two real data sets to demonstrate the effectiveness and advantages of our models in both class discovery and class prediction settings. One data set is the gene expression profiling of seventytwo acute leukemia patients [5]. In this data set, twentyfive patients were diagnosed as acute myeloid leukemia (AML) and fortyseven patients were diagnosed as acute lymphoblastic leukemia (ALL). ALL can be further divided into two groups: B cell ALLs and T cell ALLs. Totally the expressions of 6817 genes were profiled by DNA microarrays, which provide systematic information to accurately diagnose patients. The other data set is the gene expression profiling of stromal and epithelial cells of five normal and twentyeight breast cancer patients, in which the normal samples provide proper controls to highlight the specific gene expression patterns of breast cancer samples [25]. Here we apply our model (1)(2) to investigate the intrinsic structure of these samples for both class discovery and class prediction to illustrate the advantages of our model.
Leukemia data
Clustering accuracy of various methods on leukemia data
Methods  AML vs ALLs  AMLs vs B cell ALLs vs T cell ALLs 

OTCC  98%  96% 
kmeans*  98%  71% 
Spectral clustering in jClust  97%  85% 
Affinity propagation in jClust^  97%  94% 
Hierarchical clustering  98%  76% 
Applying the spectral clustering to the same similarity matrix (implemented in jClust [29]), the AMLs are grouped correctly except sample 14 and 25. This is similar to our method. But it cannot distinguish B cell ALLs from T cell ALLs (T cell ALLs merged with B cell ALLs completely). Even if changing the input similarity matrix of spectral clustering to the pairwise Pearson correlation coefficients of the gene expression profiles, spectral clustering cannot discriminate AMLs from ALLs.
We also evaluated the affinity propagation algorithm [24] implemented in jClust [29]. The affinity propagation algorithm inputs similarity scores between samples and does not require a predefined number of clusters. We find that our method outperforms jClust in accuracy using the same similarity matrix as our algorithm. In total seven groups are generated by affinity propagation with default parameters. Two groups are AMLs and other groups are ALLs. Sample 25 is misclassified as ALL whereas sample 56 is misclassified as AML. Sample 72 is misclustered with B cell ALLs and sample 35 is misclassified as T cell ALLs. Changing the input similarity matrix to the pairwise Pearson correlation coefficients of the gene expression profiles, the affinity propagation algorithm generates three groups, AMLs, B cell ALLs and T cell ALLs, but the accuracy is even lower (61/72 = 84.7%). We also tried different parameter values and the clustering accuracy cannot be further improved.
Agglomerative hierarchical clustering is another popular tool for analyzing the subtle structure underlying the gene expression profiles of cancer samples. Applying agglomerative hierarchical clustering with Euclidean distance to the AMLs and ALLs dataset, it can identify AMLs from ALLs except sample 25. But it failed to discriminate B cell ALLs from T cell ALLs (accuracy: 31/47 = 66%). The T cell ALLs and a set of sixteen B cell ALLs form one cluster whereas other B cell ALLs form the other cluster. The failure of the agglomerative hierarchical clustering for discriminating T cell ALLs from B cell ALLs can be attributed to the fact that the bottomup cluster merge strategy is a greedy one and cannot find global optimum.
Breast cancer data
Property summary of OTCC compared to other methods
Gene expression profiling technologies, e.g. microarrays and deep sequencing, have become more and more important for clinical practices, such as diagnosis and prognosis. Class discovery and class prediction are two typical tasks to utilize gene expression profiling technologies to leverage the quality and efficiency of diagnosis and prognosis. In this study, we propose a novel optimization model and integrate two tasks in one framework by treating class discovery and class prediction as a process of labeling. By seeking an optimal labeling scheme that fits best to the gene expression profiling of samples, a convex quadratic programming model is established. It can be solved efficiently and the global optimum solution is guaranteed. It does not need manual intervention to set a cutoff and can detect outliers to improve the statistical signal in the data. It does not use directly the clinical measurement but rather uses a similarity matrix as its input. The biomarker identification process is thus separated from class discovery and class prediction, facilitating clinicians to integrate prior knowledge with the measurements. It can also be applied to multiple types of measurements, e.g. gene expression profiling, proteomic analysis, and nextgeneration sequencing. Because the similarity matrix is the only input, the output is sensitive to biomarker selection and similarity measures choices. Proper biomarkers and similarity measures will generate reasonable accuracy and greatly accelerate understanding of the nature of diseases. Numerical experiments on leukemia and breast cancer data sets suggest that it is very effective for revealing and predicting the subtle subtypes of cancers based on the gene expression data of patients.
Because the objective function of our model is a quadratic form of the Laplacian matrix, it is closely related to spectral clustering and semisupervised learning methods. Spectral clustering can be generally solved by seeking the Fiedler vector of the Laplacian matrix [22, 26]. The resulting vector sums to zero and the norm equals to one. Because it originates from the matrix eigenvector, it does not provide a natural threshold. So it needs additional selection of cutoffs [22]. Here we model the class discovery and class prediction by explicitly denoting classes by zero and one and then seeking an optimal label assignment to extract the information hiding in the data. A natural cutoff, 0.5, is provided. As opposed to many semisupervised learning methods in which the unlabeled samples are assigned zero, the positive samples are assigned +1 and the negative samples are assigned −1 [23], we do not assign any labels to the unknown samples, which may prevent artificial bias during modeling. Compared to the frequently used agglomerative hierarchical clustering method, OTCC provides a divisive hierarchical clustering procedure in which the global information is utilized at each step. Compared to kmeans and fuzzy cmeans methods, OTCC can guarantee the global optimum and does not require a predefined number of clusters. This feature is helpful when clinicians do not know how many subclasses exist for a certain disease. Because it is based on the similarity matrix, it is an open framework that allows prior information to plug in. Numerical experiments on real leukemia and breast cancer data sets suggest the effectiveness of our method, especially its advantage in illustrating the fine cluster structure. Adding partial label information, OTCC turns into a class prediction tool and can reach high accuracy. We note that spectral clustering has also been extended to incorporate constraint information [30, 31] for semisupervised learning. This extends the scope of this study, so the corresponding comparison is not included.
Conclusions
Class discovery and class prediction are two tasks linked to each other inherently in clinical research. Previous studies proposed methods for these two tasks separately. And thus ignored the linkage between these two tasks. In this study, we model class discovery and class prediction in one framework and facilitate the discovery and prediction of subtle subtypes of cancers. Because of its flexibility, our method can be applied to multiple types of measurements, e.g. gene expression profiling, proteomic analysis, and nextgeneration sequencing and allows the integration of extensive prior information.
Abbreviations
 HC:

hierarchical clustering
 SOM:

selforganizing maps
 NMF:

nonnegative matrix factorization
 OTCC:

an Optimization Tool for Clustering and Classification
 SMO:

sequential minimal optimization algorithm
 AML:

acute myeloid leukemia
 ALL:

acute lymphoblastic leukemia.
Declarations
Acknowledgments
The authors thank the members of the ZHANGroup of Academy of Mathematics and Systems Science, Chinese Academy of Science for their valuable discussion and comments. This work is partially supported by the Grant No. 61171007 and No. 11131009 from the National Natural Science Foundation of China.
Authors’ Affiliations
References
 Bals R, Jany B: Identification of disease genes by expression profiling. Eur Respir J 2001, 18(5):882–889. 10.1183/09031936.01.00106601View ArticlePubMedGoogle Scholar
 Greenberg SA: DNA microarray gene expression analysis technology and its application to neurological disorders. Neurology 2001, 57(5):755–761. 10.1212/WNL.57.5.755View ArticlePubMedGoogle Scholar
 Henriksen PA, Kotelevtsev Y: Application of gene expression profiling to cardiovascular disease. Cardiovasc Res 2002, 54(1):16–24. 10.1016/S00086363(01)005168View ArticlePubMedGoogle Scholar
 Lagraulet A: Current Clinical and Pharmaceutical Applications of Microarrays: From Disease Biomarkers Discovery to Automated Diagnostics. J Assoc Lab Autom 2010, 15(5):405–413. 10.1016/j.jala.2010.06.011View ArticleGoogle Scholar
 Golub TR, Slonim DK, Tamayo P, Huard C, Gaasenbeek M, Mesirov JP, Coller H, Loh ML, Downing JR, Caligiuri MA, et al.: Molecular Classification of Cancer: Class Discovery and Class Prediction by Gene Expression Monitoring. Science 1999, 286(5439):531–537. 10.1126/science.286.5439.531View ArticlePubMedGoogle Scholar
 Brunet JP, Tamayo P, Golub TR, Mesirov JP: Metagenes and molecular pattern discovery using matrix factorization. Proc Nat Acad Sci USA 2004, 101(12):4164–4169. 10.1073/pnas.0308531101PubMed CentralView ArticlePubMedGoogle Scholar
 Gao Y, Church G: Improving molecular cancer class discovery through sparse nonnegative matrix factorization. Bioinformatics 2005, 21(21):3970–3975. 10.1093/bioinformatics/bti653View ArticlePubMedGoogle Scholar
 Hsu AL, Tang SL, Halgamuge SK: An unsupervised hierarchical dynamic selforganizing approach to cancer class discovery and marker gene identification in microarray data. Bioinformatics 2003, 19(16):2131–2140. 10.1093/bioinformatics/btg296View ArticlePubMedGoogle Scholar
 Kim H, Park H: Sparse nonnegative matrix factorizations via alternating nonnegativityconstrained least squares for microarray data analysis. Bioinformatics 2007, 23(12):1495–1502. 10.1093/bioinformatics/btm134View ArticlePubMedGoogle Scholar
 Li W, Fan M, Xiong M: SamCluster: an integrated scheme for automatic discovery of sample classes using gene expression profile. Bioinformatics 2003, 19(7):811–817. 10.1093/bioinformatics/btg095View ArticlePubMedGoogle Scholar
 Steinfeld I, Navon R, Ardigo D, Zavaroni I, Yakhini Z: Clinically driven semisupervised class discovery in gene expression data. Bioinformatics 2008, 24(16):i90i97. 10.1093/bioinformatics/btn279View ArticlePubMedGoogle Scholar
 Varma S, Simon R: Iterative class discovery and feature selection using Minimal Spanning Trees. BMC Bioinforma 2004, 5: 126. 10.1186/147121055126View ArticleGoogle Scholar
 von Heydebreck A, Huber W, Poustka A, Vingron M: Identifying splits with clear separation: a new class discovery method for gene expression data. Bioinformatics 2001, 17(suppl 1):S107S114. 10.1093/bioinformatics/17.suppl_1.S107View ArticlePubMedGoogle Scholar
 Yu Z, Wong HS, Wang H: Graphbased consensus clustering for class discovery from gene expression data. Bioinformatics 2007, 23(21):2888–2896. 10.1093/bioinformatics/btm463View ArticlePubMedGoogle Scholar
 Brown MPS, Grundy WN, Lin D, Cristianini N, Sugnet CW, Furey TS, Ares M, Haussler D: Knowledgebased analysis of microarray gene expression data by using support vector machines. ProcNat Acad Sci USA 2000, 97(1):262–267. 10.1073/pnas.97.1.262View ArticleGoogle Scholar
 Furey TS, Cristianini N, Duffy N, Bednarski DW, Schummer M, Haussler D: Support vector machine classification and validation of cancer tissue samples using microarray expression data. Bioinformatics 2000, 16(10):906–914. 10.1093/bioinformatics/16.10.906View ArticlePubMedGoogle Scholar
 Ji Y, Tsui KW, Kim K: A novel means of using gene clusters in a twostep empirical Bayes method for predicting classes of samples. Bioinformatics 2005, 21(7):1055–1061. 10.1093/bioinformatics/bti092View ArticlePubMedGoogle Scholar
 Lee Y, Lee CK: Classification of multiple cancer types by multicategory support vector machines using gene expression data. Bioinformatics 2003, 19(9):1132–1139. 10.1093/bioinformatics/btg102View ArticlePubMedGoogle Scholar
 Tan AC, Naiman DQ, Xu L, Winslow RL, Geman D: Simple decision rules for classifying human cancers from gene expression profiles. Bioinformatics 2005, 21(20):3896–3904. 10.1093/bioinformatics/bti631PubMed CentralView ArticlePubMedGoogle Scholar
 Alexandridis R, Lin S, Irwin M: Class discovery and classification of tumor samples using mixture modeling of gene expression data}a unified approach. Bioinformatics 2004, 20(16):2545–2552. 10.1093/bioinformatics/bth281View ArticlePubMedGoogle Scholar
 Filippone M, Camastra F, Masulli F, Rovetta S: Asurvey of kernel and spectral methods for clustering. Pattern Recognit 2007, 41: 176–190.View ArticleGoogle Scholar
 von Luxburg U: A Tutorial on Spectral Clustering. Stat Comput 2007, 17: 395–416. 10.1007/s112220079033zView ArticleGoogle Scholar
 Hwang T, Sicotte H, Tian Z, Wu B, Kocher JP, Wigle DA, Kumar V, Kuang R: Robust and efficient identification of biomarkers by classifying features on graphs. Bioinformatics 2008, 24(18):2023–2029. 10.1093/bioinformatics/btn383View ArticlePubMedGoogle Scholar
 Frey BJ, Dueck D: Clustering by Passing Messages Between Data Points. Science 2007, 315(5814):972–976. 10.1126/science.1136800View ArticlePubMedGoogle Scholar
 Casey T, Bond J, Tighe S, Hunter T, Lintault L, Patel O, Eneman J, Crocker A, White J, Tessitore J, et al.: Molecular signatures suggest a major role for stromal cells in development of invasive breast cancer. Breast Cancer Res Treat 2009, 114(1):47–62. 10.1007/s1054900899828View ArticlePubMedGoogle Scholar
 Kim C, Cheon M, Kang M, Chang I: A simple and exact Laplacian clustering of complex networking phenomena: Application to gene expression profiles. Proc Nat Acad Sci USA 2008, 105(11):4083–4087. 10.1073/pnas.0708598105PubMed CentralView ArticlePubMedGoogle Scholar
 Macqueen JB: Some Methods for classification and analysis of multivariate observations. In: 1967. University of California Press, Berkeley; 1967:281–297.Google Scholar
 Lloyd S: Least squares quantization in PCM. Inf Theory, IEEE Trans on 1982, 28(2):129–137. 10.1109/TIT.1982.1056489View ArticleGoogle Scholar
 Pavlopoulos GA, Moschopoulos CN, Hooper SD, Schneider R, Kossida S: jClust: A clustering and visualization toolbox. Bioinformatics 2009, 25(15):1994–1996. 10.1093/bioinformatics/btp330PubMed CentralView ArticlePubMedGoogle Scholar
 Yang C, Zhang X, Jiao L, Wang G: SelfTuning SemiSupervised Spectral Clustering. Comput Intell Secur, Int Conf on 2008, 1: 1–5.View ArticleGoogle Scholar
 Mishra A, Gillies D: Semi Supervised Spectral Clustering for Regulatory Module Discovery. In Data Integration in the Life Sciences. Edited by: Bairoch A, CohenBoulakia S, Froidevaux C. Berlin/Heidelberg, SpringerVerlag; 2008:192–203. vol. 5109 vol. 5109View ArticleGoogle Scholar
Copyright
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.