A stable iterative method for refining discriminative gene clusters

Background

Microarray technology is often used to identify the genes that are differentially expressed between two biological conditions. On the other hand, since microarray datasets contain a small number of samples and a large number of genes, it is usually desirable to identify small gene subsets with distinct pattern between sample classes. Such gene subsets are highly discriminative in phenotype classification because of their tightly coupling features. Unfortunately, such identified classifiers usually tend to have poor generalization properties on the test samples due to overfitting problem.

Results

We propose a novel approach combining both supervised learning with unsupervised learning techniques to generate increasingly discriminative gene clusters in an iterative manner. Our experiments on both simulated and real datasets show that our method can produce a series of robust gene clusters with good classification performance compared with existing approaches.

Conclusion

This backward approach for refining a series of highly discriminative gene clusters for classification purpose proves to be very consistent and stable when applied to various types of training samples.

     

Cluster-Rasch models for microarray gene expression data

Background

We propose two different formulations of the Rasch statistical models to the problem of relating gene expression profiles to the phenotypes. One formulation allows us to investigate whether a cluster of genes with similar expression profiles is related to the observed phenotypes; this model can also be used for future prediction. The other formulation provides an alternative way of identifying genes that are over- or underexpressed from their expression levels in tissue or cell samples of a given tissue or cell type.

Results

We illustrate the methods on available datasets of a classification of acute leukemias and of 60 cancer cell lines. For tumor classification, the results are comparable to those previously obtained. For the cancer cell lines dataset, we found four clusters of genes that are related to drug response for many of the 90 drugs that we considered. In addition, for each type of cell line, we identified genes that are over- or underexpressed relative to other genes.

Conclusions

The cluster-Rasch model provides a probabilistic model for describing gene expression patterns across samples and can be used to relate gene expression profiles to phenotypes.     

Learning Predictive Interactions Using Information Gain and Bayesian Network Scoring

Background

The problems of correlation and classification are long-standing in the fields of statistics and machine learning, and techniques have been developed to address these problems. We are now in the era of high-dimensional data, which is data that can concern billions of variables. These data present new challenges. In particular, it is difficult to discover predictive variables, when each variable has little marginal effect. An example concerns Genome-wide Association Studies (GWAS) datasets, which involve millions of single nucleotide polymorphism (SNPs), where some of the SNPs interact epistatically to affect disease status. Towards determining these interacting SNPs, researchers developed techniques that addressed this specific problem. However, the problem is more general, and so these techniques are applicable to other problems concerning interactions. A difficulty with many of these techniques is that they do not distinguish whether a learned interaction is actually an interaction or whether it involves several variables with strong marginal effects.

Methodology/Findings

We address this problem using information gain and Bayesian network scoring. First, we identify candidate interactions by determining whether together variables provide more information than they do separately. Then we use Bayesian network scoring to see if a candidate interaction really is a likely model. Our strategy is called MBS-IGain. Using 100 simulated datasets and a real GWAS Alzheimer’s dataset, we investigated the performance of MBS-IGain.

Conclusions/Significance

When analyzing the simulated datasets, MBS-IGain substantially out-performed nine previous methods at locating interacting predictors, and at identifying interactions exactly. When analyzing the real Alzheimer’s dataset, we obtained new results and results that substantiated previous findings. We conclude that MBS-IGain is highly effective at finding interactions in high-dimensional datasets. This result is significant because we have increasingly abundant high-dimensional data in many domains, and to learn causes and perform prediction/classification using these data, we often must first identify interactions.     

Chromothripsis is a common mechanism driving genomic rearrangements in primary and metastatic colorectal cancer

Background

Structural rearrangements form a major class of somatic variation in cancer genomes. Local chromosome shattering, termed chromothripsis, is a mechanism proposed to be the cause of clustered chromosomal rearrangements and was recently described to occur in a small percentage of tumors. The significance of these clusters for tumor development or metastatic spread is largely unclear.

Results

We used genome-wide long mate-pair sequencing and SNP array profiling to reveal that chromothripsis is a widespread phenomenon in primary colorectal cancer and metastases. We find large and small chromothripsis events in nearly every colorectal tumor sample and show that several breakpoints of chromothripsis clusters and isolated rearrangements affect cancer genes, including NOTCH2, EXO1 and MLL3. We complemented the structural variation studies by sequencing the coding regions of a cancer exome in all colorectal tumor samples and found somatic mutations in 24 genes, including APC, KRAS, SMAD4 and PIK3CA. A pairwise comparison of somatic variations in primary and metastatic samples indicated that many chromothripsis clusters, isolated rearrangements and point mutations are exclusively present in either the primary tumor or the metastasis and may affect cancer genes in a lesion-specific manner.

Conclusions

We conclude that chromothripsis is a prevalent mechanism driving structural rearrangements in colorectal cancer and show that a complex interplay between point mutations, simple copy number changes and chromothripsis events drive colorectal tumor development and metastasis.     

Biclustering via optimal re-ordering of data matrices in systems biology: rigorous methods and comparative studies

Background  

The analysis of large-scale data sets via clustering techniques is utilized in a number of applications. Biclustering in particular has emerged as an important problem in the analysis of gene expression data since genes may only jointly respond over a subset of conditions. Biclustering algorithms also have important applications in sample classification where, for instance, tissue samples can be classified as cancerous or normal. Many of the methods for biclustering, and clustering algorithms in general, utilize simplified models or heuristic strategies for identifying the "best" grouping of elements according to some metric and cluster definition and thus result in suboptimal clusters.     

A combinational feature selection and ensemble neural network method for classification of gene expression data

Background  

Microarray experiments are becoming a powerful tool for clinical diagnosis, as they have the potential to discover gene expression patterns that are characteristic for a particular disease. To date, this problem has received most attention in the context of cancer research, especially in tumor classification. Various feature selection methods and classifier design strategies also have been generally used and compared. However, most published articles on tumor classification have applied a certain technique to a certain dataset, and recently several researchers compared these techniques based on several public datasets. But, it has been verified that differently selected features reflect different aspects of the dataset and some selected features can obtain better solutions on some certain problems. At the same time, faced with a large amount of microarray data with little knowledge, it is difficult to find the intrinsic characteristics using traditional methods. In this paper, we attempt to introduce a combinational feature selection method in conjunction with ensemble neural networks to generally improve the accuracy and robustness of sample classification.     

Clustering of the SOM easily reveals distinct gene expression patterns: results of a reanalysis of lymphoma study

Background  

A method to evaluate and analyze the massive data generated by series of microarray experiments is of utmost importance to reveal the hidden patterns of gene expression. Because of the complexity and the high dimensionality of microarray gene expression profiles, the dimensional reduction of raw expression data and the feature selections necessary for, for example, classification of disease samples remains a challenge. To solve the problem we propose a two-level analysis. First self-organizing map (SOM) is used. SOM is a vector quantization method that simplifies and reduces the dimensionality of original measurements and visualizes individual tumor sample in a SOM component plane. Next, hierarchical clustering and K-means clustering is used to identify patterns of gene expression useful for classification of samples.     

MCMC implementation of the optimal Bayesian classifier for non-Gaussian models: model-based RNA-Seq classification

Background

Sequencing datasets consist of a finite number of reads which map to specific regions of a reference genome. Most effort in modeling these datasets focuses on the detection of univariate differentially expressed genes. However, for classification, we must consider multiple genes and their interactions.

Results

Thus, we introduce a hierarchical multivariate Poisson model (MP) and the associated optimal Bayesian classifier (OBC) for classifying samples using sequencing data. Lacking closed-form solutions, we employ a Monte Carlo Markov Chain (MCMC) approach to perform classification. We demonstrate superior or equivalent classification performance compared to typical classifiers for two synthetic datasets and over a range of classification problem difficulties. We also introduce the Bayesian minimum mean squared error (MMSE) conditional error estimator and demonstrate its computation over the feature space. In addition, we demonstrate superior or leading class performance over an RNA-Seq dataset containing two lung cancer tumor types from The Cancer Genome Atlas (TCGA).

Conclusions

Through model-based, optimal Bayesian classification, we demonstrate superior classification performance for both synthetic and real RNA-Seq datasets. A tutorial video and Python source code is available under an open source license at http://bit.ly/1gimnss.

Electronic supplementary material

The online version of this article (doi:10.1186/s12859-014-0401-3) contains supplementary material, which is available to authorized users.     

Increasing the Number of Thyroid Lesions Classes in Microarray Analysis Improves the Relevance of Diagnostic Markers

Background

Genetic markers for thyroid cancers identified by microarray analysis have offered limited predictive accuracy so far because of the few classes of thyroid lesions usually taken into account. To improve diagnostic relevance, we have simultaneously analyzed microarray data from six public datasets covering a total of 347 thyroid tissue samples representing 12 histological classes of follicular lesions and normal thyroid tissue. Our own dataset, containing about half the thyroid tissue samples, included all categories of thyroid lesions.

Methodology/Principal Findings

Classifier predictions were strongly affected by similarities between classes and by the number of classes in the training sets. In each dataset, sample prediction was improved by separating the samples into three groups according to class similarities. The cross-validation of differential genes revealed four clusters with functional enrichments. The analysis of six of these genes (APOD, APOE, CLGN, CRABP1, SDHA and TIMP1) in 49 new samples showed consistent gene and protein profiles with the class similarities observed. Focusing on four subclasses of follicular tumor, we explored the diagnostic potential of 12 selected markers (CASP10, CDH16, CLGN, CRABP1, HMGB2, ALPL2, ADAMTS2, CABIN1, ALDH1A3, USP13, NR2F2, KRTHB5) by real-time quantitative RT-PCR on 32 other new samples. The gene expression profiles of follicular tumors were examined with reference to the mutational status of the Pax8-PPARγ, TSHR, GNAS and NRAS genes.

Conclusion/Significance

We show that diagnostic tools defined on the basis of microarray data are more relevant when a large number of samples and tissue classes are used. Taking into account the relationships between the thyroid tumor pathologies, together with the main biological functions and pathways involved, improved the diagnostic accuracy of the samples. Our approach was particularly relevant for the classification of microfollicular adenomas.     

Distinct Transmission Networks of Chlamydia trachomatis in Men Who Have Sex with Men and Heterosexual Adults in Amsterdam,The Netherlands

Background

Genovar distributions of Chlamydia trachomatis based on ompA typing differ between men who have sex with men (MSM) and heterosexuals. We investigated clonal relationships using a high resolution typing method to characterize C. trachomatis types in these two risk groups.

Methods

C. trachomatis positive samples were collected at the STI outpatient clinic in Amsterdam between 2008 and 2010 and genotyped by multilocus sequence typing. Clusters were assigned using minimum spanning trees and these were combined with epidemiological data of the hosts.

Results

We typed 526 C. trachomatis positive samples: 270 from MSM and 256 from heterosexuals. Eight clusters, containing 10–128 samples were identified of which 4 consisted of samples from MSM (90%–100%), with genovars D, G, J, and L2b. The other 4 clusters consisted mainly of samples from heterosexuals (87%–100%) with genovars D, E, F, I, and J. Genetic diversity was much lower in the MSM clusters than in heterosexual clusters. Significant differences in number of sexual partners and HIV-serostatus were observed for MSM–associated clusters.

Conclusions

C. trachomatis transmission patterns among MSM and heterosexuals were largely distinct. We hypothesize that these differences are due to sexual host behavior, but bacterial factors may play a role as well.     

Analyzing the similarity of samples and genes by MG-PCC algorithm,t-SNE-SS and t-SNE-SG maps

Background

For analyzing these gene expression data sets under different samples, clustering and visualizing samples and genes are important methods. However, it is difficult to integrate clustering and visualizing techniques when the similarities of samples and genes are defined by PCC(Person correlation coefficient) measure.

Results

Here, for rare samples of gene expression data sets, we use MG-PCC (mini-groups that are defined by PCC) algorithm to divide them into mini-groups, and use t-SNE-SSP maps to display these mini-groups, where the idea of MG-PCC algorithm is that the nearest neighbors should be in the same mini-groups, t-SNE-SSP map is selected from a series of t-SNE(t-statistic Stochastic Neighbor Embedding) maps of standardized samples, and these t-SNE maps have different perplexity parameter. Moreover, for PCC clusters of mass genes, they are displayed by t-SNE-SGI map, where t-SNE-SGI map is selected from a series of t-SNE maps of standardized genes, and these t-SNE maps have different initialization dimensions. Here, t-SNE-SSP and t-SNE-SGI maps are selected by A-value, where A-value is modeled from areas of clustering projections, and t-SNE-SSP and t-SNE-SGI maps are such t-SNE map that has the smallest A-value.

Conclusions

From the analysis of cancer gene expression data sets, we demonstrate that MG-PCC algorithm is able to put tumor and normal samples into their respective mini-groups, and t-SNE-SSP(or t-SNE-SGI) maps are able to display the relationships between mini-groups(or PCC clusters) clearly. Furthermore, t-SNE-SS(m)(or t-SNE-SG(n)) maps are able to construct independent tree diagrams of the nearest sample(or gene) neighbors, where each tree diagram is corresponding to a mini-group of samples(or genes).

     

Defining adult asthma endotypes by clinical features and patterns of volatile organic compounds in exhaled air

Background

Several classifications of adult asthma patients using cluster analyses based on clinical and demographic information has resulted in clinical phenotypic clusters that do not address molecular mechanisms. Volatile organic compounds (VOC) in exhaled air are released during inflammation in response to oxidative stress as a result of activated leukocytes. VOC profiles in exhaled air could distinguish between asthma patients and healthy subjects. In this study, we aimed to classify new asthma endotypes by combining inflammatory mechanisms investigated by VOC profiles in exhaled air and clinical information of asthma patients.

Methods

Breath samples were analyzed for VOC profiles by gas chromatography–mass spectrometry from asthma patients (n = 195) and healthy controls (n = 40). A total of 945 determined compounds were subjected to discriminant analysis to find those that could discriminate healthy from asthmatic subjects. 2-step cluster analysis based on clinical information and VOCs in exhaled air were used to form asthma endotypes.

Results

We identified 16 VOCs, which could distinguish between healthy and asthma subjects with a sensitivity of 100% and a specificity of 91.1%. Cluster analysis based on VOCs in exhaled air and the clinical parameters FEV1, FEV1 change after 3 weeks of hospitalization, allergic sensitization, Junipers symptoms score and asthma medications resulted in the formation of 7 different asthma endotype clusters. We identified asthma clusters with different VOC profiles but similar clinical characteristics and endotypes with similar VOC profiles, but distinct clinical characteristics.

Conclusion

This study demonstrates that both, clinical presentation of asthma and inflammatory mechanisms in the airways should be considered for classification of asthma subtypes.

Electronic supplementary material

The online version of this article (doi:10.1186/s12931-014-0136-8) contains supplementary material, which is available to authorized users.     

Evaluating somatic tumor mutation detection without matched normal samples

Background

Observations of recurrent somatic mutations in tumors have led to identification and definition of signaling and other pathways that are important for cancer progression and therapeutic targeting. As tumor cells contain both an individual’s inherited genetic variants and somatic mutations, challenges arise in distinguishing these events in massively parallel sequencing datasets. Typically, both a tumor sample and a “normal” sample from the same individual are sequenced and compared; variants observed only in the tumor are considered to be somatic mutations. However, this approach requires two samples for each individual.

Results

We evaluate a method of detecting somatic mutations in tumor samples for which only a subset of normal samples are available. We describe tuning of the method for detection of mutations in tumors, filtering to remove inherited variants, and comparison of detected mutations to several matched tumor/normal analysis methods. Filtering steps include the use of population variation datasets to remove inherited variants as well a subset of normal samples to remove technical artifacts. We then directly compare mutation detection with tumor-only and tumor-normal approaches using the same sets of samples. Comparisons are performed using an internal targeted gene sequencing dataset (n = 3380) as well as whole exome sequencing data from The Cancer Genome Atlas project (n = 250). Tumor-only mutation detection shows similar recall (43–60%) but lesser precision (20–21%) to current matched tumor/normal approaches (recall 43–73%, precision 30–82%) when compared to a “gold-standard” tumor/normal approach. The inclusion of a small pool of normal samples improves precision, although many variants are still uniquely detected in the tumor-only analysis.

Conclusions

A detailed method for somatic mutation detection without matched normal samples enables study of larger numbers of tumor samples, as well as tumor samples for which a matched normal is not available. As sensitivity/recall is similar to tumor/normal mutation detection but precision is lower, tumor-only detection is more appropriate for classification of samples based on known mutations. Although matched tumor-normal analysis is preferred due to higher precision, we demonstrate that mutation detection without matched normal samples is possible for certain applications.

     

Quantifying stability in gene list ranking across microarray derived clinical biomarkers

Background

Genomic instability in cancer leads to abnormal genome copy number alterations (CNA) as a mechanism underlying tumorigenesis. Using microarrays and other technologies, tumor CNA are detected by comparing tumor sample CN to normal reference sample CN. While advances in microarray technology have improved detection of copy number alterations, the increase in the number of measured signals, noise from array probes, variations in signal-to-noise ratio across batches and disparity across laboratories leads to significant limitations for the accurate identification of CNA regions when comparing tumor and normal samples.

Methods

To address these limitations, we designed a novel "Virtual Normal" algorithm (VN), which allowed for construction of an unbiased reference signal directly from test samples within an experiment using any publicly available normal reference set as a baseline thus eliminating the need for an in-lab normal reference set.

Results

The algorithm was tested using an optimal, paired tumor/normal data set as well as previously uncharacterized pediatric malignant gliomas for which a normal reference set was not available. Using Affymetrix 250K Sty microarrays, we demonstrated improved signal-to-noise ratio and detected significant copy number alterations using the VN algorithm that were validated by independent PCR analysis of the target CNA regions.

Conclusions

We developed and validated an algorithm to provide a virtual normal reference signal directly from tumor samples and minimize noise in the derivation of the raw CN signal. The algorithm reduces the variability of assays performed across different reagent and array batches, methods of sample preservation, multiple personnel, and among different laboratories. This approach may be valuable when matched normal samples are unavailable or the paired normal specimens have been subjected to variations in methods of preservation.     

PureCN: copy number calling and SNV classification using targeted short read sequencing

Background

Matched sequencing of both tumor and normal tissue is routinely used to classify variants of uncertain significance (VUS) into somatic vs. germline. However, assays used in molecular diagnostics focus on known somatic alterations in cancer genes and often only sequence tumors. Therefore, an algorithm that reliably classifies variants would be helpful for retrospective exploratory analyses. Contamination of tumor samples with normal cells results in differences in expected allelic fractions of germline and somatic variants, which can be exploited to accurately infer genotypes after adjusting for local copy number. However, existing algorithms for determining tumor purity, ploidy and copy number are not designed for unmatched short read sequencing data.

Results

We describe a methodology and corresponding open source software for estimating tumor purity, copy number, loss of heterozygosity (LOH), and contamination, and for classification of single nucleotide variants (SNVs) by somatic status and clonality. This R package, PureCN, is optimized for targeted short read sequencing data, integrates well with standard somatic variant detection pipelines, and has support for matched and unmatched tumor samples. Accuracy is demonstrated on simulated data and on real whole exome sequencing data.

Conclusions

Our algorithm provides accurate estimates of tumor purity and ploidy, even if matched normal samples are not available. This in turn allows accurate classification of SNVs. The software is provided as open source (Artistic License 2.0) R/Bioconductor package PureCN (http://bioconductor.org/packages/PureCN/).

     

Biogeography at the limits of life: Do extremophilic microbial communities show biogeographical regionalization?

Aim

Biogeographical regions are the fundamental geographical units for grouping Earth's biodiversity. Biogeographical regionalization has been demonstrated for many higher taxa, such as terrestrial plants and vertebrates, but not in microbial communities. Therefore, we sought to test empirically whether microbial communities, or taxa, show patterns consistent with biogeographical regionalization.

Location

Within halite (NaCl) crystals from coastal solar salterns of western Europe, the Mediterranean and east Africa.

Time period

Modern (2006–2013).

Major taxa studied

Archaea.

Methods

Using high‐throughput Illumina amplicon sequencing, we generated the most high‐resolution characterization of halite‐associated archaeal communities to date, using samples from 17 locations. We grouped communities into biogeographical clusters based on community turnover to test whether these communities show biogeographical regionalization. To examine whether individual taxa, rather than communities, show biogeographical patterns, we also tested whether the relative abundance of individual genera may be indicative of a community's biogeographical origins using machine learning methods, specifically random forest classification.

Results

We found that the rate of community turnover was greatest over subregional spatial scales (< 500 km), whereas at regional spatial scales the turnover was independent of geographical distance. Biogeographical clusters of communities were either not statistically robust or lacked spatial coherence, inconsistent with biogeographical regionalization. However, we identified several archaeal genera that were good indicators of biogeographical origin, providing classification error rates of < 10%.

Main conclusions

Overall, our results provide little support for the concept of biogeographical regions in these extremophilic microbial communities, despite the fact that some taxa do show biogeographical patterns. We suggest that variable dispersal ability among the halite‐associated Archaea may disrupt biogeographical patterns at the community level, preventing the formation of biogeographical regions. This means that the processes that lead to the formation of biogeographical regions operate differentially on individual microbial taxa rather than on entire communities.     

Degenerative Pathways of Lumbar Motion Segments - A Comparison in Two Samples of Patients with Persistent Low Back Pain

Background

Magnetic resonance imaging (MRI) is used to identify spinal pathoanatomy in people with persistent low back pain. However, the clinical relevance of spinal degenerative MRI findings remains uncertain. Although multiple MRI findings are almost always present at the same time, research into the association with clinical outcomes (such as pain) has predominantly focused on individual MRI findings. This study aimed to: (i) investigate how multiple MRI lumbar spine findings cluster together within two different samples of patients with low back pain, (ii) classify these clusters into hypothetical pathways of degeneration based on scientific knowledge of disco-vertebral degeneration, and (iii) compare these clusters and degenerative pathways between samples.

Methods

We performed a secondary cross-sectional analysis on two dissimilar MRI samples collected in a hospital department: (1) data from the spinal MRI reports of 4,162 low back pain patients and (2) data from an MRI research protocol of 631 low back pain patients. Latent Class Analysis was used in both samples to cluster MRI findings from lumbar motion segments. Using content analysis, each cluster was then categorised into hypothetical pathways of degeneration.

Results

Six clusters of MRI findings were identified in each of the two samples. The content of the clusters in the two samples displayed some differences but had the same overall pattern of MRI findings. Although the hypothetical degenerative pathways identified in the two samples were not identical, the overall pattern of increasing degeneration within the pathways was the same.

Conclusions

It was expected that different clusters could emerge from different samples, however, when organised into hypothetical pathways of degeneration, the overall pattern of increasing degeneration was similar and biologically plausible. This evidence of reproducibility suggests that Latent Class Analysis may provide a new approach to investigating the relationship between MRI findings and clinically important characteristics such as pain and activity limitation.     

Diffusion Model Based Spectral Clustering for Protein-Protein Interaction Networks

Background

A goal of systems biology is to analyze large-scale molecular networks including gene expressions and protein-protein interactions, revealing the relationships between network structures and their biological functions. Dividing a protein-protein interaction (PPI) network into naturally grouped parts is an essential way to investigate the relationship between topology of networks and their functions. However, clear modular decomposition is often hard due to the heterogeneous or scale-free properties of PPI networks.

Methodology/Principal Findings

To address this problem, we propose a diffusion model-based spectral clustering algorithm, which analytically solves the cluster structure of PPI networks as a problem of random walks in the diffusion process in them. To cope with the heterogeneity of the networks, the power factor is introduced to adjust the diffusion matrix by weighting the transition (adjacency) matrix according to a node degree matrix. This algorithm is named adjustable diffusion matrix-based spectral clustering (ADMSC). To demonstrate the feasibility of ADMSC, we apply it to decomposition of a yeast PPI network, identifying biologically significant clusters with approximately equal size. Compared with other established algorithms, ADMSC facilitates clear and fast decomposition of PPI networks.

Conclusions/Significance

ADMSC is proposed by introducing the power factor that adjusts the diffusion matrix to the heterogeneity of the PPI networks. ADMSC effectively partitions PPI networks into biologically significant clusters with almost equal sizes, while being very fast, robust and appealing simple.     

Accurate molecular classification of cancer using simple rules

Background

One intractable problem with using microarray data analysis for cancer classification is how to reduce the extremely high-dimensionality gene feature data to remove the effects of noise. Feature selection is often used to address this problem by selecting informative genes from among thousands or tens of thousands of genes. However, most of the existing methods of microarray-based cancer classification utilize too many genes to achieve accurate classification, which often hampers the interpretability of the models. For a better understanding of the classification results, it is desirable to develop simpler rule-based models with as few marker genes as possible.

Methods

We screened a small number of informative single genes and gene pairs on the basis of their depended degrees proposed in rough sets. Applying the decision rules induced by the selected genes or gene pairs, we constructed cancer classifiers. We tested the efficacy of the classifiers by leave-one-out cross-validation (LOOCV) of training sets and classification of independent test sets.

Results

We applied our methods to five cancerous gene expression datasets: leukemia (acute lymphoblastic leukemia [ALL] vs. acute myeloid leukemia [AML]), lung cancer, prostate cancer, breast cancer, and leukemia (ALL vs. mixed-lineage leukemia [MLL] vs. AML). Accurate classification outcomes were obtained by utilizing just one or two genes. Some genes that correlated closely with the pathogenesis of relevant cancers were identified. In terms of both classification performance and algorithm simplicity, our approach outperformed or at least matched existing methods.

Conclusion

In cancerous gene expression datasets, a small number of genes, even one or two if selected correctly, is capable of achieving an ideal cancer classification effect. This finding also means that very simple rules may perform well for cancerous class prediction.