Multiclass cancer classification and biomarker discovery using GA-based algorithms

A Unified Multi-Class Feature Selection Framework for Microarray Data

In feature selection research, simultaneous multi-class feature selection technologies are popular because they simultaneously select informative features for all classes. Recursive feature elimination (RFE) methods are state-of-the-art binary feature selection algorithms. However, extending existing RFE algorithms to multi-class tasks may increase the computational cost and lead to performance degradation. With this motivation, we introduce a unified multi-class feature selection (UFS) framework for randomization-based neural networks to address these challenges. First, we propose a new multi-class feature ranking criterion using the output weights of neural networks. The heuristic underlying this criterion is that "the importance of a feature should be related to the magnitude of the output weights of a neural network". Subsequently, the UFS framework utilizes the original features to construct a training model based on a randomization-based neural network, ranks these features by the criterion of the norm of the output weights, and recursively removes a feature with the lowest ranking score. Extensive experiments on 15 real-world datasets suggest that our proposed framework outperforms state-of-the-art algorithms. The code of UFS is available at https://github.com/SVMrelated/UFS.git.

Tumor type classification and candidate cancer-specific biomarkers discovery via semi-supervised learning

Identifying cancer-related differentially expressed genes provides significant information for diagnosing tumors, predicting prognoses, and effective treatments. Recently, deep learning methods have been used to perform gene differential expression analysis using microarray-based high-throughput gene profiling and have achieved good results. In this study, we proposed a new robust multiple-datasets-based semi-supervised learning model, MSSL, to perform tumor type classification and candidate cancer-specific biomarkers discovery across multiple tumor types and multiple datasets, which addressed the following long-lasting obstacles: (1) the data volume of the existing single dataset is not enough to fully exert the advantages of deep learning; (2) a large number of datasets from different research institutions cannot be effectively used due to inconsistent internal variances and low quality; (3) relatively uncommon cancers have limited effects on deep learning methods. In our article, we applied MSSL to The Cancer Genome Atlas (TCGA) and the Gene Expression Comprehensive Database (GEO) pan-cancer normalized-level3 RNA-seq data and got 97.6% final classification accuracy, which had a significant performance leap compared with previous approaches. Finally, we got the ranking of the importance of the corresponding genes for each cancer type based on classification results and validated that the top genes selected in this way were biologically meaningful for corresponding tumors and some of them had been used as biomarkers, which showed the efficacy of our method.

Cancer classification based on multiple dimensions: SNV patterns

BACKGROUND

The occurrence of cancer is closely related to single nucleotide variants (SNVs). However, in DNA samples collected from patients with distinct cancers, SNVs are detected in different patterns. Therefore, it is an important task to select the appropriate method by which to classify cancer to the greatest extent of SNV patterns, which will aid in cancer diagnosis and treatment. In traditional studies, researchers combined each SNV with its neighboring nucleotides to form a trinucleotide. Mutation signatures for cancer classification were extracted from the patterns of the trinucleotides, but the SNV feature extraction in a single dimension may result in partial information loss and poor model performance.

RESULTS

In this study, we defined multidimensional SNV (M-SNV) features to classify cancer. M-SNV features considered first- and second-order neighboring nucleotides of one-dimensional SNVs and included six types of features. We validated the feasibility of M-SNV features using a dataset obtained from The Cancer Genome Atlas (TCGA) consisting of 2761 samples from 12 cancers. We performed preliminary screening of 562,321 DNA mutation sites in these samples. The remaining mutation sites were characterized by cancer type in six signatures. We found that the extracted features showed a similar distribution in the cluster center of the cancer type of the samples. After the preprocessing of raw data, samples were more focused on the cancer subtype distributions at the SNV level. We used KNN (k-nearest neighbors) to classify the extracted features and employed the leave-one-out cross to verify them. The accuracy of classifying is stable at approximately 97% and can reach 97.43% in the most optimal case. Furthermore, we found that the validated oncogenes in the loci of the features had the highest importance among the 8 cancers.

CONCLUSIONS

It is feasible to classify cancers by the distribution of features we defined. Moreover, our methodology has potential implications for the discovery of oncogenes.

Defining and predicting transdiagnostic categories of neurodegenerative disease

The prevalence of concomitant proteinopathies and heterogeneous clinical symptoms in neurodegenerative diseases hinders the identification of individuals who might be candidates for a particular intervention. Here, by applying an unsupervised clustering algorithm to post-mortem histopathological data from 895 patients with degeneration in the central nervous system, we show that six non-overlapping disease clusters can simultaneously account for tau neurofibrillary tangles, α-synuclein inclusions, neuritic plaques, inclusions of the transcriptional repressor TDP-43, angiopathy, neuron loss and gliosis. We also show that membership to the six transdiagnostic disease clusters, which explains more variance in cognitive phenotypes than can be explained by individual diagnoses, can be accurately predicted from scores of the Mini-Mental Status Exam, protein levels in cerebrospinal fluid, and genotype at the APOE and MAPT loci, via cross-validated multiple logistic regression. This combination of unsupervised and supervised data-driven tools provides a framework that could be used to identify latent disease subtypes in other areas of medicine.

Classification of Cancer Types Using Graph Convolutional Neural Networks

BACKGROUND

Cancer has been a leading cause of death in the United States with significant health care costs. Accurate prediction of cancers at an early stage and understanding the genomic mechanisms that drive cancer development are vital to the improvement of treatment outcomes and survival rates, thus resulting in significant social and economic impacts. Attempts have been made to classify cancer types with machine learning techniques during the past two decades and deep learning approaches more recently.

RESULTS

In this paper, we established four models with graph convolutional neural network (GCNN) that use unstructured gene expressions as inputs to classify different tumor and non-tumor samples into their designated 33 cancer types or as normal. Four GCNN models based on a co-expression graph, co-expression+singleton graph, protein-protein interaction (PPI) graph, and PPI+singleton graph have been designed and implemented. They were trained and tested on combined 10,340 cancer samples and 731 normal tissue samples from The Cancer Genome Atlas (TCGA) dataset. The established GCNN models achieved excellent prediction accuracies (89.9-94.7%) among 34 classes (33 cancer types and a normal group). In silico gene-perturbation experiments were performed on four models based on co-expression graph, co-expression+singleton, PPI graph, and PPI+singleton graphs. The co-expression GCNN model was further interpreted to identify a total of 428 markers genes that drive the classification of 33 cancer types and normal. The concordance of differential expressions of these markers between the represented cancer type and others are confirmed. Successful classification of cancer types and a normal group regardless of normal tissues' origin suggested that the identified markers are cancer-specific rather than tissue-specific.

CONCLUSION

Novel GCNN models have been established to predict cancer types or normal tissue based on gene expression profiles. We demonstrated the results from the TCGA dataset that these models can produce accurate classification (above 94%), using cancer-specific markers genes. The models and the source codes are publicly available and can be readily adapted to the diagnosis of cancer and other diseases by the data-driven modeling research community.

Bounded Fuzzy Possibilistic Method Reveals Information about Lung Cancer through Analysis of Metabolomics

Learning methods, such as conventional clustering and classification, have been applied in diagnosing diseases to categorize samples based on their features. Going beyond clustering samples, membership degrees represent to what degree each sample belongs to a cluster. Variation of membership degrees in each cluster provides information about the cluster as a whole and each sample individually which enables us to have insights toward precision medicine. Membership degrees are measured more accurately through removing restrictions from clustering samples. Bounded Fuzzy Possibilistic Method (BFPM) introduces a membership function that keeps the search space flexible to cluster samples with higher accuracy. The method evaluates samples for their movement from one cluster to another. This technique allows us to find critical samples in advance those with the potential ability to belong to other clusters in the near future. BFPM was applied on metabolomics of individuals in a lung cancer case-control study. Metabolomics as proximal molecular signals to the actual disease processes may serve as strong biomarkers of current disease process. The goal is to know whether serum metabolites of a healthy human can be differentiated from those with lung cancer. Using BFPM, some differences were observed, the pathology data were evaluated, and critical samples were recognized.

A comparative study of multi-omics integration tools for cancer driver gene identification and tumour subtyping

Oncogenesis and cancer can arise as a consequence of a wide range of genomic aberrations including mutations, copy number alterations, expression changes and epigenetic modifications encompassing multiple omics layers. Integrating genomic, transcriptomic, proteomic and epigenomic datasets via multi-omics analysis provides the opportunity to derive a deeper and holistic understanding of the development and progression of cancer. There are two primary approaches to integrating multi-omics data: multi-staged (focused on identifying genes driving cancer) and meta-dimensional (focused on establishing clinically relevant tumour or sample classifications). A number of ready-to-use bioinformatics tools are available to perform both multi-staged and meta-dimensional integration of multi-omics data. In this study, we compared nine different integration tools using real and simulated cancer datasets. The performance of the multi-staged integration tools were assessed at the gene, function and pathway levels, while meta-dimensional integration tools were assessed based on the sample classification performance. Additionally, we discuss the influence of factors such as data representation, sample size, signal and noise on multi-omics data integration. Our results provide current and much needed guidance regarding selection and use of the most appropriate and best performing multi-omics integration tools.

Deciphering essential cistromes using genome-wide CRISPR screens

Systematically dissecting the function of a large set of cis-regulatory elements or transcription factor binding sites (cistromes) has been technically challenging. Using genome-wide CRISPR screens, we profiled over 10,000 FOXA1 and CTCF binding sites for their roles in regulating the fitness of breast and prostate cancer cells, and accordingly developed a model to predict essentiality for cis-elements. These efforts not only reveal how the key transcription factors and their cistromes regulate cell essentiality in hormone-dependent cancers but also highlight an efficient approach to investigate the functions of noncoding regions of the genome.

Although millions of transcription factor binding sites, or cistromes, have been identified across the human genome, defining which of these sites is functional in a given condition remains challenging. Using CRISPR/Cas9 knockout screens and gene essentiality or fitness as the readout, we systematically investigated the essentiality of over 10,000 FOXA1 and CTCF binding sites in breast and prostate cancer cells. We found that essential FOXA1 binding sites act as enhancers to orchestrate the expression of nearby essential genes through the binding of lineage-specific transcription factors. In contrast, CRISPR screens of the CTCF cistrome revealed 2 classes of essential binding sites. The first class of essential CTCF binding sites act like FOXA1 sites as enhancers to regulate the expression of nearby essential genes, while a second class of essential CTCF binding sites was identified at topologically associated domain (TAD) boundaries and display distinct characteristics. Using regression methods trained on our screening data and public epigenetic profiles, we developed a model to predict essential cis-elements with high accuracy. The model for FOXA1 essentiality correctly predicts noncoding variants associated with cancer risk and progression. Taken together, CRISPR screens of cis-regulatory elements can define the essential cistrome of a given factor and can inform the development of predictive models of cistrome function.

Differential gene expression analysis reveals novel genes and pathways in pediatric septic shock patients

Septic shock is a devastating health condition caused by uncontrolled sepsis. Advancements in high-throughput sequencing techniques have increased the number of potential genetic biomarkers under review. Multiple genetic markers and functional pathways play a part in development and progression of pediatric septic shock. We identified 53 differentially expressed pediatric septic shock biomarkers using gene expression data sampled from 181 patients admitted to the pediatric intensive care unit within the first 24 hours of their admission. The gene expression signatures showed discriminatory power between pediatric septic shock survivors and nonsurvivor types. Using functional enrichment analysis of differentially expressed genes, we validated the known genes and pathways in septic shock and identified the unexplored septic shock-related genes and functional groups. Differential gene expression analysis revealed the genes involved in the immune response, chemokine-mediated signaling, neutrophil chemotaxis, and chemokine activity and distinguished the septic shock survivor from non-survivor. The identification of the septic shock gene biomarkers may facilitate in septic shock diagnosis, treatment, and prognosis.

Discovering cooperative biomarkers for heterogeneous complex disease diagnoses

Biomarkers with high reproducibility and accurate prediction performance can contribute to comprehending the underlying pathogenesis of related complex diseases and further facilitate disease diagnosis and therapy. Techniques integrating gene expression profiles and biological networks for the identification of network-based disease biomarkers are receiving increasing interest. The biomarkers for heterogeneous diseases often exhibit strong cooperative effects, which implies that a set of genes may achieve more accurate outcome prediction than any single gene. In this study, we evaluated various biomarker identification methods that consider gene cooperative effects implicitly or explicitly, and proposed the gene cooperation network to explicitly model the cooperative effects of gene combinations. The gene cooperation network-enhanced method, named as MarkRank, achieves superior performance compared with traditional biomarker identification methods in both simulation studies and real data sets. The biomarkers identified by MarkRank not only have a better prediction accuracy but also have stronger topological relationships in the biological network and exhibit high specificity associated with the related diseases. Furthermore, the top genes identified by MarkRank involve crucial biological processes of related diseases and give a good prioritization for known disease genes. In conclusion, MarkRank suggests that explicit modeling of gene cooperative effects can greatly improve biomarker identification for complex diseases, especially for diseases with high heterogeneity.

Tumor classification and biomarker discovery based on the 5'isomiR expression level

Background

The miRNA isoforms (isomiRs) have been suggested to regulate the same pathways as the canonical miRNA and play an important biological role in miRNA-mediated gene regulation. Recently, a study has demonstrated that the presence or absence of all isomiRs could efficiently discriminate amongst 32 TCGA cancer types. Besides, an effective reduction of distinguishing isomiR features for multiclass tumor discrimination must have a major impact on our understanding of the disease and treatment of cancer.

Methods

In this study, we have constructed a combination of the genetic algorithms (GA) with Random Forest (RF) algorithms to detect reliable sets of cancer-associated 5’isomiRs from TCGA isomiR expression data for multiclass tumor classification.

Results

We obtained 100 sets of the optimal predictive features, each of which comprised of 50–5’isomiRs that could effectively classify with an average sensitivity of 92% samples from 32 different tumor types. We calculated the frequency with which a 5’isomiR found in these sets as measuring its importance for tumor classification. Many highly frequent 5’isomiRs with different 5′ loci from canonical miRNAs were detected in these sets, supporting that the isomiRs play a significant role in the multiclass tumor classification. The further functional enrichment analysis showed that the target genes of the 10 most frequently appearing 5’isomiRs were involved in the activity of transcription activator and protein kinase and cell-cell adhesion.

Conclusions

The findings of the present study indicated that the 5’isomiRs might be employed for multiclass tumor classification and the suggested that GA/RF model could perform effective tumor classification by a series of largely independent optimal predictor 5′ isomiR sets.

A data-driven, knowledge-based approach to biomarker discovery: application to circulating microRNA markers of colorectal cancer prognosis

Recent advances in high-throughput technologies have provided an unprecedented opportunity to identify molecular markers of disease processes. This plethora of complex-omics data has simultaneously complicated the problem of extracting meaningful molecular signatures and opened up new opportunities for more sophisticated integrative and holistic approaches. In this era, effective integration of data-driven and knowledge-based approaches for biomarker identification has been recognised as key to improving the identification of high-performance biomarkers, and necessary for translational applications. Here, we have evaluated the role of circulating microRNA as a means of predicting the prognosis of patients with colorectal cancer, which is the second leading cause of cancer-related death worldwide. We have developed a multi-objective optimisation method that effectively integrates a data-driven approach with the knowledge obtained from the microRNA-mediated regulatory network to identify robust plasma microRNA signatures which are reliable in terms of predictive power as well as functional relevance. The proposed multi-objective framework has the capacity to adjust for conflicting biomarker objectives and to incorporate heterogeneous information facilitating systems approaches to biomarker discovery. We have found a prognostic signature of colorectal cancer comprising 11 circulating microRNAs. The identified signature predicts the patients' survival outcome and targets pathways underlying colorectal cancer progression. The altered expression of the identified microRNAs was confirmed in an independent public data set of plasma samples of patients in early stage vs advanced colorectal cancer. Furthermore, the generality of the proposed method was demonstrated across three publicly available miRNA data sets associated with biomarker studies in other diseases.

An Empirical Study of Univariate and Genetic Algorithm-Based Feature Selection in Binary Classification with Microarray Data

Background

We consider both univariate- and multivariate-based feature selection for the problem of binary classification with microarray data. The idea is to determine whether the more sophisticated multivariate approach leads to better misclassification error rates because of the potential to consider jointly significant subsets of genes (but without overfitting the data).

Methods

We present an empirical study in which 10-fold cross-validation is applied externally to both a univariate-based and two multivariate- (genetic algorithm (GA)-) based feature selection processes. These procedures are applied with respect to three supervised learning algorithms and six published two-class microarray datasets.

Results

Considering all datasets, and learning algorithms, the average 10-fold external cross-validation error rates for the univariate-, single-stage GA-, and two-stage GA-based processes are 14.2%, 14.6%, and 14.2%, respectively. We also find that the optimism bias estimates from the GA analyses were half that of the univariate approach, but the selection bias estimates from the GA analyses were 2.5 times that of the univariate results.

Conclusions

We find that the 10-fold external cross-validation misclassification error rates were very comparable. Further, we find that a two-stage GA approach did not demonstrate a significant advantage over a 1-stage approach. We also find that the univariate approach had higher optimism bias and lower selection bias compared to both GA approaches.

Cross-platform Analysis of Cancer Biomarkers: A Bayesian Network Approach to Incorporating Mass Spectrometry and Microarray Data

Many studies showed inconsistent cancer biomarkers due to bioinformatics artifacts. In this paper we use multiple data sets from microarrays, mass spectrometry, protein sequences, and other biological knowledge in order to improve the reliability of cancer biomarkers. We present a novel Bayesian network (BN) model which integrates and cross-annotates multiple data sets related to prostate cancer. The main contribution of this study is that we provide a method that is designed to find cancer biomarkers whose presence is supported by multiple data sources and biological knowledge. Relevant biological knowledge is explicitly encoded into the model parameters, and the biomarker finding problem is formulated as a Bayesian inference problem. Besides diagnostic accuracy, we introduce reliability as another quality measurement of the biological relevance of biomarkers. Based on the proposed BN model, we develop an empirical scoring scheme and a simulation algorithm for inferring biomarkers. Fourteen genes/proteins including prostate specific antigen (PSA) are identified as reliable serum biomarkers which are insensitive to the model assumptions. The computational results show that our method is able to find biologically relevant biomarkers with highest reliability while maintaining competitive predictive power. In addition, by combining biological knowledge and data from multiple platforms, the number of putative biomarkers is greatly reduced to allow more-focused clinical studies.

Identification of potential tissue-specific cancer biomarkers and development of cancer versus normal genomic classifiers

Machine learning techniques for cancer prediction and biomarker discovery can hasten cancer detection and significantly improve prognosis. Recent “OMICS” studies which include a variety of cancer and normal tissue samples along with machine learning approaches have the potential to further accelerate such discovery. To demonstrate this potential, 2,175 gene expression samples from nine tissue types were obtained to identify gene sets whose expression is characteristic of each cancer class. Using random forests classification and ten-fold cross-validation, we developed nine single-tissue classifiers, two multi-tissue cancer-versus-normal classifiers, and one multi-tissue normal classifier. Given a sample of a specified tissue type, the single-tissue models classified samples as cancer or normal with a testing accuracy between 85.29% and 100%. Given a sample of non-specific tissue type, the multi-tissue bi-class model classified the sample as cancer versus normal with a testing accuracy of 97.89%. Given a sample of non-specific tissue type, the multi-tissue multi-class model classified the sample as cancer versus normal and as a specific tissue type with a testing accuracy of 97.43%. Given a normal sample of any of the nine tissue types, the multi-tissue normal model classified the sample as a particular tissue type with a testing accuracy of 97.35%. The machine learning classifiers developed in this study identify potential cancer biomarkers with sensitivity and specificity that exceed those of existing biomarkers and pointed to pathways that are critical to tissue-specific tumor development. This study demonstrates the feasibility of predicting the tissue origin of carcinoma in the context of multiple cancer classes.

Early response index: a statistic to discover potential early stage disease biomarkers

Background

Identifying disease correlated features early before large number of molecules are impacted by disease progression with significant abundance change is very advantageous to biologists for developing early disease diagnosis biomarkers. Disease correlated features have relatively low level of abundance change at early stages. Finding them using existing bioinformatic tools in high throughput data is a challenging task since the technology suffers from limited dynamic range and significant noise. Most existing biomarker discovery algorithms can only detect molecules with high abundance changes, frequently missing early disease diagnostic markers.

Results

We present a new statistic called early response index (ERI) to prioritize disease correlated molecules as potential early biomarkers. Instead of classification accuracy, ERI measures the average classification accuracy improvement attainable by a feature when it is united with other counterparts for classification. ERI is more sensitive to abundance changes than other ranking statistics. We have shown that ERI significantly outperforms SAM and Localfdr in detecting early responding molecules in a proteomics study of a mouse model of multiple sclerosis. Importantly, ERI was able to detect many disease relevant proteins before those algorithms detect them at a later time point.

Conclusions

ERI method is more sensitive for significant feature detection during early stage of disease development. It potentially has a higher specificity for biomarker discovery, and can be used to identify critical time frame for disease intervention.

Electronic supplementary material

The online version of this article (doi:10.1186/s12859-017-1712-y) contains supplementary material, which is available to authorized users.

The parameter sensitivity of random forests

Background

The Random Forest (RF) algorithm for supervised machine learning is an ensemble learning method widely used in science and many other fields. Its popularity has been increasing, but relatively few studies address the parameter selection process: a critical step in model fitting. Due to numerous assertions regarding the performance reliability of the default parameters, many RF models are fit using these values. However there has not yet been a thorough examination of the parameter-sensitivity of RFs in computational genomic studies. We address this gap here.

Results

We examined the effects of parameter selection on classification performance using the RF machine learning algorithm on two biological datasets with distinct p/n ratios: sequencing summary statistics (low p/n) and microarray-derived data (high p/n). Here, p, refers to the number of variables and, n, the number of samples. Our findings demonstrate that parameterization is highly correlated with prediction accuracy and variable importance measures (VIMs). Further, we demonstrate that different parameters are critical in tuning different datasets, and that parameter-optimization significantly enhances upon the default parameters.

Conclusions

Parameter performance demonstrated wide variability on both low and high p/n data. Therefore, there is significant benefit to be gained by model tuning RFs away from their default parameter settings.

Electronic supplementary material

The online version of this article (doi:10.1186/s12859-016-1228-x) contains supplementary material, which is available to authorized users.

Gene selection for cancer classification with the help of bees

BACKGROUND

Development of biologically relevant models from gene expression data notably, microarray data has become a topic of great interest in the field of bioinformatics and clinical genetics and oncology. Only a small number of gene expression data compared to the total number of genes explored possess a significant correlation with a certain phenotype. Gene selection enables researchers to obtain substantial insight into the genetic nature of the disease and the mechanisms responsible for it. Besides improvement of the performance of cancer classification, it can also cut down the time and cost of medical diagnoses.

METHODS

This study presents a modified Artificial Bee Colony Algorithm (ABC) to select minimum number of genes that are deemed to be significant for cancer along with improvement of predictive accuracy. The search equation of ABC is believed to be good at exploration but poor at exploitation. To overcome this limitation we have modified the ABC algorithm by incorporating the concept of pheromones which is one of the major components of Ant Colony Optimization (ACO) algorithm and a new operation in which successive bees communicate to share their findings.

RESULTS

The proposed algorithm is evaluated using a suite of ten publicly available datasets after the parameters are tuned scientifically with one of the datasets. Obtained results are compared to other works that used the same datasets. The performance of the proposed method is proved to be superior.

CONCLUSION

The method presented in this paper can provide subset of genes leading to more accurate classification results while the number of selected genes is smaller. Additionally, the proposed modified Artificial Bee Colony Algorithm could conceivably be applied to problems in other areas as well.

Helminths and the microbiota: parts of the hygiene hypothesis

In modern societies, diseases that are driven by dysregulated immune responses are increasing at an alarming pace, such as inflammatory bowel diseases and diabetes. There is an urgent need to understand these epidemiological trends, which are likely to be driven by the changing environment of the last few decades. There are complex interactions between human genetic factors and this changing environment that is leading to the increasing prevalence of metabolic and inflammatory diseases. Alterations to human gut bacterial communities (the microbiota) and lowered prevalence of helminth infections are potential environmental factors contributing to immune dysregulation. Helminths have co-evolved with the gut microbiota and their mammalian hosts. This three-way interaction is beginning to be characterized, and the knowledge gained may enable the design of new therapeutic strategies to treat metabolic and inflammatory diseases. However, these complex interactions need to be carefully investigated in the context of host genetic backgrounds to identify optimal treatment strategies. The complex nature of these interactions raises the possibility that only with highly personalized treatment, with knowledge of individual genetic and microbiota communities, will therapeutic interventions be successful for a majority of the individuals suffering from these complex diseases of immune dysregulation.

Environmental mold and mycotoxin exposures elicit specific cytokine and chemokine responses

Background

Molds can cause respiratory symptoms and asthma. We sought to use isolated peripheral blood mononuclear cells (PBMCs) to understand changes in cytokine and chemokine levels in response to mold and mycotoxin exposures and to link these levels with respiratory symptoms in humans. We did this by utilizing an ex vivo assay approach to differentiate mold-exposed patients and unexposed controls. While circulating plasma chemokine and cytokine levels from these two groups might be similar, we hypothesized that by challenging their isolated white blood cells with mold or mold extracts, we would see a differential chemokine and cytokine release.

Methods and Findings

Peripheral blood mononuclear cells (PBMCs) were isolated from blood from 33 patients with a history of mold exposures and from 17 controls. Cultured PBMCs were incubated with the most prominent Stachybotrys chartarum mycotoxin, satratoxin G, or with aqueous mold extract, ionomycin, or media, each with or without PMA. Additional PBMCs were exposed to spores of Aspergillus niger, Cladosporium herbarum and Penicillium chrysogenum. After 18 hours, cytokines and chemokines released into the culture medium were measured by multiplex assay. Clinical histories, physical examinations and pulmonary function tests were also conducted. After ex vivo PBMC exposures to molds or mycotoxins, the chemokine and cytokine profiles from patients with a history of mold exposure were significantly different from those of unexposed controls. In contrast, biomarker profiles from cells exposed to media alone showed no difference between the patients and controls.

Conclusions

These findings demonstrate that chronic mold exposures induced changes in inflammatory and immune system responses to specific mold and mycotoxin challenges. These responses can differentiate mold-exposed patients from unexposed controls. This strategy may be a powerful approach to document immune system responsiveness to molds and other inflammation-inducing environmental agents.

Multiple biomarker panels for early detection of breast cancer in peripheral blood

Detecting breast cancer at early stages can be challenging. Traditional mammography and tissue microarray that have been studied for early breast cancer detection and prediction have many drawbacks. Therefore, there is a need for more reliable diagnostic tools for early detection of breast cancer due to a number of factors and challenges. In the paper, we presented a five-marker panel approach based on SVM for early detection of breast cancer in peripheral blood and show how to use SVM to model the classification and prediction problem of early detection of breast cancer in peripheral blood. We found that the five-marker panel can improve the prediction performance (area under curve) in the testing data set from 0.5826 to 0.7879. Further pathway analysis showed that the top four five-marker panels are associated with signaling, steroid hormones, metabolism, immune system, and hemostasis, which are consistent with previous findings. Our prediction model can serve as a general model for multibiomarker panel discovery in early detection of other cancers.

A classification framework applied to cancer gene expression profiles

Classification of cancer based on gene expression has provided insight into possible treatment strategies. Thus, developing machine learning methods that can successfully distinguish among cancer subtypes or normal versus cancer samples is important. This work discusses supervised learning techniques that have been employed to classify cancers. Furthermore, a two-step feature selection method based on an attribute estimation method (e.g., ReliefF) and a genetic algorithm was employed to find a set of genes that can best differentiate between cancer subtypes or normal versus cancer samples. The application of different classification methods (e.g., decision tree, k-nearest neighbor, support vector machine (SVM), bagging, and random forest) on 5 cancer datasets shows that no classification method universally outperforms all the others. However, k-nearest neighbor and linear SVM generally improve the classification performance over other classifiers. Finally, incorporating diverse types of genomic data (e.g., protein-protein interaction data and gene expression) increase the prediction accuracy as compared to using gene expression alone.

Early diagnosis of complex diseases by molecular biomarkers, network biomarkers, and dynamical network biomarkers

Many studies have been carried out for early diagnosis of complex diseases by finding accurate and robust biomarkers specific to respective diseases. In particular, recent rapid advance of high-throughput technologies provides unprecedented rich information to characterize various disease genotypes and phenotypes in a global and also dynamical manner, which significantly accelerates the study of biomarkers from both theoretical and clinical perspectives. Traditionally, molecular biomarkers that distinguish disease samples from normal samples are widely adopted in clinical practices due to their ease of data measurement. However, many of them suffer from low coverage and high false-positive rates or high false-negative rates, which seriously limit their further clinical applications. To overcome those difficulties, network biomarkers (or module biomarkers) attract much attention and also achieve better performance because a network (or subnetwork) is considered to be a more robust form to characterize diseases than individual molecules. But, both molecular biomarkers and network biomarkers mainly distinguish disease samples from normal samples, and they generally cannot ensure to identify predisease samples due to their static nature, thereby lacking ability to early diagnosis. Based on nonlinear dynamical theory and complex network theory, a new concept of dynamical network biomarkers (DNBs, or a dynamical network of biomarkers) has been developed, which is different from traditional static approaches, and the DNB is able to distinguish a predisease state from normal and disease states by even a small number of samples, and therefore has great potential to achieve "real" early diagnosis of complex diseases. In this paper, we comprehensively review the recent advances and developments on molecular biomarkers, network biomarkers, and DNBs in particular, focusing on the biomarkers for early diagnosis of complex diseases considering a small number of samples and high-throughput data (or big data). Detailed comparisons of various types of biomarkers as well as their applications are also discussed.

Evaluation of breast cancer susceptibility using improved genetic algorithms to generate genotype SNP barcodes

Genetic association is a challenging task for the identification and characterization of genes that increase the susceptibility to common complex multifactorial diseases. To fully execute genetic studies of complex diseases, modern geneticists face the challenge of detecting interactions between loci. A genetic algorithm (GA) is developed to detect the association of genotype frequencies of cancer cases and noncancer cases based on statistical analysis. An improved genetic algorithm (IGA) is proposed to improve the reliability of the GA method for high-dimensional SNP-SNP interactions. The strategy offers the top five results to the random population process, in which they guide the GA toward a significant search course. The IGA increases the likelihood of quickly detecting the maximum ratio difference between cancer cases and noncancer cases. The study systematically evaluates the joint effect of 23 SNP combinations of six steroid hormone metabolisms, and signaling-related genes involved in breast carcinogenesis pathways were systematically evaluated, with IGA successfully detecting significant ratio differences between breast cancer cases and noncancer cases. The possible breast cancer risks were subsequently analyzed by odds-ratio (OR) and risk-ratio analysis. The estimated OR of the best SNP barcode is significantly higher than 1 (between 1.15 and 7.01) for specific combinations of two to 13 SNPs. Analysis results support that the IGA provides higher ratio difference values than the GA between breast cancer cases and noncancer cases over 3-SNP to 13-SNP interactions. A more specific SNP-SNP interaction profile for the risk of breast cancer is also provided.

Finding minimum gene subsets with heuristic breadth-first search algorithm for robust tumor classification

Background

Previous studies on tumor classification based on gene expression profiles suggest that gene selection plays a key role in improving the classification performance. Moreover, finding important tumor-related genes with the highest accuracy is a very important task because these genes might serve as tumor biomarkers, which is of great benefit to not only tumor molecular diagnosis but also drug development.

Results

This paper proposes a novel gene selection method with rich biomedical meaning based on Heuristic Breadth-first Search Algorithm (HBSA) to find as many optimal gene subsets as possible. Due to the curse of dimensionality, this type of method could suffer from over-fitting and selection bias problems. To address these potential problems, a HBSA-based ensemble classifier is constructed using majority voting strategy from individual classifiers constructed by the selected gene subsets, and a novel HBSA-based gene ranking method is designed to find important tumor-related genes by measuring the significance of genes using their occurrence frequencies in the selected gene subsets. The experimental results on nine tumor datasets including three pairs of cross-platform datasets indicate that the proposed method can not only obtain better generalization performance but also find many important tumor-related genes.

Conclusions

It is found that the frequencies of the selected genes follow a power-law distribution, indicating that only a few top-ranked genes can be used as potential diagnosis biomarkers. Moreover, the top-ranked genes leading to very high prediction accuracy are closely related to specific tumor subtype and even hub genes. Compared with other related methods, the proposed method can achieve higher prediction accuracy with fewer genes. Moreover, they are further justified by analyzing the top-ranked genes in the context of individual gene function, biological pathway, and protein-protein interaction network.

Learning biomarkers of pluripotent stem cells in mouse

Pluripotent stem cells are able to self-renew, and to differentiate into all adult cell types. Many studies report data describing these cells, and characterize them in molecular terms. Machine learning yields classifiers that can accurately identify pluripotent stem cells, but there is a lack of studies yielding minimal sets of best biomarkers (genes/features). We assembled gene expression data of pluripotent stem cells and non-pluripotent cells from the mouse. After normalization and filtering, we applied machine learning, classifying samples into pluripotent and non-pluripotent with high cross-validated accuracy. Furthermore, to identify minimal sets of best biomarkers, we used three methods: information gain, random forests and a wrapper of genetic algorithm and support vector machine (GA/SVM). We demonstrate that the GA/SVM biomarkers work best in combination with each other; pathway and enrichment analyses show that they cover the widest variety of processes implicated in pluripotency. The GA/SVM wrapper yields best biomarkers, no matter which classification method is used. The consensus best biomarker based on the three methods is Tet1, implicated in pluripotency just recently. The best biomarker based on the GA/SVM wrapper approach alone is Fam134b, possibly a missing link between pluripotency and some standard surface markers of unknown function processed by the Golgi apparatus.

A comparison of machine learning techniques for survival prediction in breast cancer

Background

The ability to accurately classify cancer patients into risk classes, i.e. to predict the outcome of the pathology on an individual basis, is a key ingredient in making therapeutic decisions. In recent years gene expression data have been successfully used to complement the clinical and histological criteria traditionally used in such prediction. Many "gene expression signatures" have been developed, i.e. sets of genes whose expression values in a tumor can be used to predict the outcome of the pathology. Here we investigate the use of several machine learning techniques to classify breast cancer patients using one of such signatures, the well established 70-gene signature.

Results

We show that Genetic Programming performs significantly better than Support Vector Machines, Multilayered Perceptrons and Random Forests in classifying patients from the NKI breast cancer dataset, and comparably to the scoring-based method originally proposed by the authors of the 70-gene signature. Furthermore, Genetic Programming is able to perform an automatic feature selection.

Conclusions

Since the performance of Genetic Programming is likely to be improvable compared to the out-of-the-box approach used here, and given the biological insight potentially provided by the Genetic Programming solutions, we conclude that Genetic Programming methods are worth further investigation as a tool for cancer patient classification based on gene expression data.

ICGA-PSO-ELM approach for accurate multiclass cancer classification resulting in reduced gene sets in which genes encoding secreted proteins are highly represented

A combination of Integer-Coded Genetic Algorithm (ICGA) and Particle Swarm Optimization (PSO), coupled with the neural-network-based Extreme Learning Machine (ELM), is used for gene selection and cancer classification. ICGA is used with PSO-ELM to select an optimal set of genes, which is then used to build a classifier to develop an algorithm (ICGA_PSO_ELM) that can handle sparse data and sample imbalance. We evaluate the performance of ICGA-PSO-ELM and compare our results with existing methods in the literature. An investigation into the functions of the selected genes, using a systems biology approach, revealed that many of the identified genes are involved in cell signaling and proliferation. An analysis of these gene sets shows a larger representation of genes that encode secreted proteins than found in randomly selected gene sets. Secreted proteins constitute a major means by which cells interact with their surroundings. Mounting biological evidence has identified the tumor microenvironment as a critical factor that determines tumor survival and growth. Thus, the genes identified by this study that encode secreted proteins might provide important insights to the nature of the critical biological features in the microenvironment of each tumor type that allow these cells to thrive and proliferate.

PUGSVM: a caBIG™ analytical tool for multiclass gene selection and predictive classification

Phenotypic Up-regulated Gene Support Vector Machine (PUGSVM) is a cancer Biomedical Informatics Grid (caBIG™) analytical tool for multiclass gene selection and classification. PUGSVM addresses the problem of imbalanced class separability, small sample size and high gene space dimensionality, where multiclass gene markers are defined by the union of one-versus-everyone phenotypic upregulated genes, and used by a well-matched one-versus-rest support vector machine. PUGSVM provides a simple yet more accurate strategy to identify statistically reproducible mechanistic marker genes for characterization of heterogeneous diseases.

AVAILABILITY

http://www.cbil.ece.vt.edu/caBIG-PUGSVM.htm.

The genetics of pain: implications for evaluation and treatment of spinal disease

BACKGROUND CONTEXT

Variability in human pain experience appears to be at least partially determined by genetic inheritance. To the extent that awareness of individual pain sensitivity and the tendency to develop chronic pain after injury or surgery would be informative for clinical decision making, development and use of genetic testing for specific pain markers could contribute to improved outcomes in management of spinal disease.

PURPOSE

To review important and illustrative results from both classical and modern pain genetics studies and to introduce readers to critical definitions and concepts necessary to interpret the growing body of genetics literature relevant to spinal disease.

STUDY DESIGN/SETTING

Literature review and commentary.

METHODS

A review was performed of published English language studies in which genetic techniques were used to analyze the molecular basis of nociceptive signaling or processing with a particular emphasis on studies addressing genetic determinants of interindividual variability in pain sensitivity or predisposition to chronic pain.

RESULTS

There is compelling evidence indicating that interindividual differences in pain sensitivity and the risk of developing chronic pain syndromes are genetically determined. Despite a growing list of putative "pain genes," genetic association studies remain plagued with difficulty replicating initial findings in different cohorts.

CONCLUSIONS

Genome-wide association studies are potentially powerful means of identifying clinically relevant genetic markers predicting disease susceptibility, severity, and treatment response. However, accurate results require rigorous study design with use of large homogeneous populations and precise phenotypes.

Neighborhood rough set reduction-based gene selection and prioritization for gene expression profile analysis and molecular cancer classification

Selection of reliable cancer biomarkers is crucial for gene expression profile-based precise diagnosis of cancer type and successful treatment. However, current studies are confronted with overfitting and dimensionality curse in tumor classification and false positives in the identification of cancer biomarkers. Here, we developed a novel gene-ranking method based on neighborhood rough set reduction for molecular cancer classification based on gene expression profile. Comparison with other methods such as PAM, ClaNC, Kruskal-Wallis rank sum test, and Relief-F, our method shows that only few top-ranked genes could achieve higher tumor classification accuracy. Moreover, although the selected genes are not typical of known oncogenes, they are found to play a crucial role in the occurrence of tumor through searching the scientific literature and analyzing protein interaction partners, which may be used as candidate cancer biomarkers.