Chromosomal microarray analysis in fetuses with high-risk prenatal indications: A retrospective study in China.

OBJECTIVE: The present study aimed to determine the diagnostic value of prenatal chromosomal microarray analysis (CMA) for fetuses with several indications of being at high risk for various conditions. MATERIALS AND METHODS: This retrospective analysis included 1256 pregnancies that were prenatally evaluated due to high-risk indications using invasive CMA. The indications for invasive prenatal diagnosis mainly included ultrasound anomalies, high-risk for maternal serum screening (MSS), high-risk for non-invasive prenatal tests (NIPT), family history of genetic disorders or birth defects, and advanced maternal age (AMA). The rate of clinically significant genomic imbalances between the different groups was compared. RESULTS: The overall prenatal diagnostic yield was 98 (7.8%) of 1256 pregnancies. Clinically significant genomic aberrations were identified in 2 (1.5%) of 132 patients with non-structural ultrasound anomalies, 36 (12.7%) of 283 with structural ultrasound anomalies, 2 (4.5%) of 44 at high-risk for MSS, 38 (26.6%) of 143 at high-risk for NIPT, 11 (3.8%) of 288 with a family history, and 7 (2.1%) of 328 with AMA. Submicroscopic findings were identified in 29 fetuses, 19 of whom showed structural ultrasound anomalies. CONCLUSION: The diagnostic yields of CMA for pregnancies with different indications greatly varied. CMA could serve as a first-tier test for structural anomalies, especially multiple anomalies, craniofacial dysplasia, urinary defects, and cardiac dysplasia. Our results have important implications for genetic counseling.

Microarray Data Mining and Preliminary Bioinformatics Analysis of Hepatitis D Virus-Associated Hepatocellular Carcinoma.

Several studies have demonstrated that chronic hepatitis delta virus (HDV) infection is associated with a worsening of hepatitis B virus (HBV) infection and increased risk of hepatocellular carcinoma (HCC). However, there is limited data on the role of HDV in the oncogenesis of HCC. This study is aimed at assessing the potential mechanisms of HDV-associated hepatocarcinogenesis, especially to screen and identify key genes and pathways possibly involved in the pathogenesis of HCC. We selected three microarray datasets: GSE55092 contains 39 cancer specimens and 81 paracancer specimens from 11 HBV-associated HCC patients, GSE98383 contains 11 cancer specimens and 24 paracancer specimens from 5 HDV-associated HCC patients, and 371 HCC patients with the RNA-sequencing data combined with their clinical data from the Cancer Genome Atlas (TCGA). Afterwards, 948 differentially expressed genes (DEGs) closely related to HDV-associated HCC were obtained using the R package and filtering with a Venn diagram. We then performed gene ontology (GO) annotation and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis to determine the biological processes (BP), cellular component (CC), molecular function (MF), and KEGG signaling pathways most enriched for DEGs. Additionally, we performed Weighted Gene Coexpression Network Analysis (WGCNA) and protein-to-protein interaction (PPI) network construction with 948 DEGs, from which one module was identified by WGCNA and three modules were identified by the PPI network. Subsequently, we validated the expression of 52 hub genes from the PPI network with an independent set of HCC dataset stored in the Gene Expression Profiling Interactive Analysis (GEPIA) database. Finally, seven potential key genes were identified by intersecting with key modules from WGCNA, including 3 reported genes, namely, CDCA5, CENPH, and MCM7, and 4 novel genes, namely, CDC6, CDC45, CDCA8, and MCM4, which are associated with nucleoplasm, cell cycle, DNA replication, and mitotic cell cycle. The CDCA8 and stage of HCC were the independent factors associated with overall survival of HDV-associated HCC. All the related findings of these genes can help gain a better understanding of the role of HDV in the underlying mechanism of HCC carcinogenesis.

A Set of Markers Related to Viral Infection Has a Sex-sensitive Prognostic Value in Papillary Thyroid Carcinoma.

CONTEXT: Although the incidence of papillary thyroid carcinoma (PTC) is significantly higher in females than in males, the prognosis of male PTC is more unfavorable. However, the cause of higher malignancy of PTC in male patients remains unclear. OBJECTIVE: We conducted our analysis on microarrays datasets, tissue samples from PTC patients and the RNAseq datasets from TCGA with survival data. METHODS: We searched all publicly available microarray datasets and performed a genome-wide meta-analysis comparing PTC and normal samples. Gene Ontology analysis was then conducted. The candidate genes were tested by quantitative real-time polymerase chain reaction. The analysis of prognostic value of genes was performed with datasets from The Cancer Genome Atlas. RESULTS: After meta-analyses, 150 significantly differentially expressed genes (DEGs) were specifically found in male subjects. Gene Ontology analysis of these 150 genes revealed that the viral process was activated. Seven genes involved in the viral process in male patients showed a significantly differential expression between PTC and normal tissue. Survival analysis exhibited that the 7 genes, used in combination, were prognostically valuable and, of them, PSMB1 possessed a conspicuous prognostic value, especially in males. CONCLUSION: In this study, we searched all publicly available microarray datasets and conducted a comprehensive analysis to understand the male propensity for higher malignancy. We found that markers of viral infection showed significantly differential expression only in male patients compared with their female counterparts and had a sex-sensitive prognostic value in PTC.

Use of microarrays and MLPA for integrating diagnostics and personalizing treatment - Case report of a patient with Ph-like acute B-cell lymphoblastic leukemia.

B-cell precursor acute lymphoblastic leukemia (BCP-ALL) is the most common childhood cancer. A special subtype of high risk BCP-ALL is Philadelphia-like ALL (Ph-like ALL), in which the gene expression profile is similar to BCR-ABL1-positive leukemia; however, fusion of the mentioned genes does not occur. The unfavourable clinical course and incidence of 15% of cases means that the diagnosis and therapeutic strategy of Ph-like ALL must be carefully developed and implemented into clinical practice. The study presents the case of a patient with diagnosed Ph-like ALL. The use of molecular analytical techniques has made it possible to identify a patient who is likely to relapse and who may benefit from personalized therapy This study shows the advantages of using genomic analyses to identify therapeutic targets, which is especially important for patients with high-risk disease. This model of management could be extended to other cancer subtypes, allowing for tailored diagnosis.

Simulation-selection-extrapolation: Estimation in high-dimensional errors-in-variables models.

Errors-in-variables models in high-dimensional settings pose two challenges in application. First, the number of observed covariates is larger than the sample size, while only a small number of covariates are true predictors under an assumption of model sparsity. Second, the presence of measurement error can result in severely biased parameter estimates, and also affects the ability of penalized methods such as the lasso to recover the true sparsity pattern. A new estimation procedure called SIMulation-SELection-EXtrapolation (SIMSELEX) is proposed. This procedure makes double use of lasso methodology. First, the lasso is used to estimate sparse solutions in the simulation step, after which a group lasso is implemented to do variable selection. The SIMSELEX estimator is shown to perform well in variable selection, and has significantly lower estimation error than naive estimators that ignore measurement error. SIMSELEX can be applied in a variety of errors-in-variables settings, including linear models, generalized linear models, and Cox survival models. It is furthermore shown in the Supporting Information how SIMSELEX can be applied to spline-based regression models. A simulation study is conducted to compare the SIMSELEX estimators to existing methods in the linear and logistic model settings, and to evaluate performance compared to naive methods in the Cox and spline models. Finally, the method is used to analyze a microarray dataset that contains gene expression measurements of favorable histology Wilms tumors.

Integrated analysis of 34 microarray datasets reveals CBX3 as a diagnostic and prognostic biomarker in glioblastoma.

BACKGROUND: Glioblastomas have a high degree of malignancy, high recurrence rate, high mortality rate, and low cure rate. Searching for new markers of glioblastomas is of great significance for improving the diagnosis, prognosis and treatment of glioma. METHODS: Using the GEO public database, we combined 34 glioma microarray datasets containing 1893 glioma samples and conducted genetic data mining through statistical analysis, bioclustering, and pathway analysis. The results were validated in TCGA, CGGA, and internal cohorts. We further selected a gene for subsequent experiments and conducted cell proliferation and cell cycle analyses to verify the biological function of this gene. RESULTS: Eight glioblastoma-specific differentially expressed genes were screened using GEO. In the TCGA and CGGA cohorts, patients with high CBX3, BARD1, EGFR, or IFRD1 expression had significantly shorter survival but patients with high GUCY1A3 or MOBP expression had significantly longer survival than patients with lower expression of these genes. After reviewing the literature, we selected the CBX3 gene for further experiments. We confirmed that CBX3 was overexpressed in glioblastoma by immunohistochemical analysis of tissue microarrays and qPCR analysis of surgical specimens. The functional assay results showed that silencing CBX3 arrests the cell cycle in the G2/M phase, thereby weakening the cell proliferation ability. CONCLUSIONS: We used a multidisciplinary approach to analyze glioblastoma samples in 34 microarray datasets, revealing novel diagnostic and prognostic biomarkers in patients with glioblastoma and providing a new direction for screening tumor markers.

Gene selection for microarray data classification via adaptive hypergraph embedded dictionary learning.

Due to the rapid development of DNA microarray technology, a large number of microarray data come into being and classifying these data has been verified useful for cancer diagnosis, treatment and prevention. However, microarray data classification is still a challenging task since there are often a huge number of genes but a small number of samples in gene expression data. As a result, a computational method for reducing the dimension of microarray data is necessary. In this paper, we introduce a computational gene selection model for microarray data classification via adaptive hypergraph embedded dictionary learning (AHEDL). Specifically, a dictionary is learned from the feature space of original high dimensional microarray data, and this learned dictionary is used to represent original genes with a reconstruction coefficient matrix. Then we use a l2, 1-norm regularization to impose the row sparsity on the coefficient matrix for selecting discriminate genes. Meanwhile, in order to capture the localmanifold geometrical structure of original microarray data in a high-order manner, a hypergraph is adaptively learned and embedded into the model. An iterative updating algorithm is designed for solving the optimization problem. In order to validate the efficacy of the proposed model, we have conducted experiments on six publicly available microarray data sets and the results demonstrate that AHEDL outperforms other state-of-the-art methods in terms of microarray data classification. ABBREVIATIONS.

Hybrid Feature Selection Algorithm mRMR-ICA for Cancer Classification from Microarray Gene Expression Data.

AIMS AND OBJECTIVE: Redundant information of microarray gene expression data makes it difficult for cancer classification. Hence, it is very important for researchers to find appropriate ways to select informative genes for better identification of cancer. This study was undertaken to present a hybrid feature selection method mRMR-ICA which combines minimum redundancy maximum relevance (mRMR) with imperialist competition algorithm (ICA) for cancer classification in this paper. MATERIALS AND METHODS: The presented algorithm mRMR-ICA utilizes mRMR to delete redundant genes as preprocessing and provide the small datasets for ICA for feature selection. It will use support vector machine (SVM) to evaluate the classification accuracy for feature genes. The fitness function includes classification accuracy and the number of selected genes. RESULTS: Ten benchmark microarray gene expression datasets are used to test the performance of mRMR-ICA. Experimental results including the accuracy of cancer classification and the number of informative genes are improved for mRMR-ICA compared with the original ICA and other evolutionary algorithms. CONCLUSION: The comparison results demonstrate that mRMR-ICA can effectively delete redundant genes to ensure that the algorithm selects fewer informative genes to get better classification results. It also can shorten calculation time and improve efficiency.

Concordance analysis of microarray studies identifies representative gene expression changes in Parkinson's disease: a comparison of 33 human and animal studies.

BACKGROUND: As the popularity of transcriptomic analysis has grown, the reported lack of concordance between different studies of the same condition has become a growing concern, raising questions as to the representativeness of different study types, such as non-human disease models or studies of surrogate tissues, to gene expression in the human condition. METHODS: In a comparison of 33 microarray studies of Parkinson's disease, correlation and clustering analyses were used to determine the factors influencing concordance between studies, including agreement between different tissue types, different microarray platforms, and between neurotoxic and genetic disease models and human Parkinson's disease. RESULTS: Concordance over all studies is low, with correlation of only 0.05 between differential gene expression signatures on average, but increases within human patients and studies of the same tissue type, rising to 0.38 for studies of human substantia nigra. Agreement of animal models, however, is dependent on model type. Studies of brain tissue from Parkinson's disease patients (specifically the substantia nigra) form a distinct group, showing patterns of differential gene expression noticeably different from that in non-brain tissues and animal models of Parkinson's disease; while comparison with other brain diseases (Alzheimer's disease and brain cancer) suggests that the mixed study types display a general signal of neurodegenerative disease. A meta-analysis of these 33 microarray studies demonstrates the greater ability of studies in humans and highly-affected tissues to identify genes previously known to be associated with Parkinson's disease. CONCLUSIONS: The observed clustering and concordance results suggest the existence of a 'characteristic' signal of Parkinson's disease found in significantly affected human tissues in humans. These results help to account for the consistency (or lack thereof) so far observed in microarray studies of Parkinson's disease, and act as a guide to the selection of transcriptomic studies most representative of the underlying gene expression changes in the human disease.

A Severe Asthma Disease Signature from Gene Expression Profiling of Peripheral Blood from U-BIOPRED Cohorts.

RATIONALE: Stratification of asthma at the molecular level, especially using accessible biospecimens, could greatly enable patient selection for targeted therapy. OBJECTIVES: To determine the value of blood analysis to identify transcriptional differences between clinically defined asthma and nonasthma groups, identify potential patient subgroups based on gene expression, and explore biological pathways associated with identified differences. METHODS: Transcriptomic profiles were generated by microarray analysis of blood from 610 patients with asthma and control participants in the U-BIOPRED (Unbiased Biomarkers in Prediction of Respiratory Disease Outcomes) study. Differentially expressed genes (DEGs) were identified by analysis of variance, including covariates for RNA quality, sex, and clinical site, and Ingenuity Pathway Analysis was applied. Patient subgroups based on DEGs were created by hierarchical clustering and topological data analysis. MEASUREMENTS AND MAIN RESULTS: A total of 1,693 genes were differentially expressed between patients with severe asthma and participants without asthma. The differences from participants without asthma in the nonsmoking severe asthma and mild/moderate asthma subgroups were significantly related (r = 0.76), with a larger effect size in the severe asthma group. The majority of, but not all, differences were explained by differences in circulating immune cell populations. Pathway analysis showed an increase in chemotaxis, migration, and myeloid cell trafficking in patients with severe asthma, decreased B-lymphocyte development and hematopoietic progenitor cells, and lymphoid organ hypoplasia. Cluster analysis of DEGs led to the creation of subgroups among the patients with severe asthma who differed in molecular responses to oral corticosteroids. CONCLUSIONS: Blood gene expression differences between clinically defined subgroups of patients with asthma and individuals without asthma, as well as subgroups of patients with severe asthma defined by transcript profiles, show the value of blood analysis in stratifying patients with asthma and identifying molecular pathways for further study. Clinical trial registered with www.clinicaltrials.gov (NCT01982162).

Chromosomal microarray use among women undergoing invasive prenatal diagnosis.

OBJECTIVE: To study the offer and uptake of chromosomal microarray analysis (CMA) among women undergoing invasive prenatal testing. METHODS: This is a retrospective cohort study of women who underwent chorionic villus sampling (CVS) or amniocentesis. Charts were reviewed for CMA offer and uptake, in additional to clinical and demographic variables. RESULTS: One hundred forty-one women underwent CVS (n = 53) or amniocentesis (n = 91) over the study period. Overall, 41% of women underwent CMA. Women who underwent invasive testing for a fetal structural abnormality were more likely to undergo CMA than women who underwent invasive testing for all other indications (aOR 43.18, 95% CI 4.64 - 401.58). Chromosomal microarray was offered more often to women who primarily spoke English (p < 0.001), self-identified as white (p = 0.046) and did not receive prenatal care in a community health center (p = 0.044). Statistically significant differences in CMA uptake by race/ethnicity, language, insurance or provider type were not noted. Multiparous women were less likely to accept this test than nulliparas (aOR 0.39, 95% CI 0.17 - 0.86). CONCLUSION: Women who undergo invasive fetal testing are more likely to undergo CMA if the indication is for a fetal structural anomaly. There may be important demographic disparities in the offering of CMA which bear further exploration. © 2016 John Wiley & Sons, Ltd.

Graph-based unsupervised feature selection and multiview clustering for microarray data.

A challenge in bioinformatics is to analyse volumes of gene expression data generated through microarray experiments and obtain useful information. Consequently, most microarray studies demand complex data analysis to infer biologically meaningful information from such high-throughput data. Selection of informative genes is an important data analysis step to identify a set of genes which can further help in finding the biological information embedded in microarray data, and thus assists in diagnosis, prognosis and treatment of the disease. In this article we present an unsupervised feature selection technique which attempts to address the goal of explorative data analysis, unfolding the multi-faceted nature of data. It focuses on extracting multiple clustering views considering the diversity of each view from high-dimensional data. We evaluated our technique on benchmark data sets and the experimental results indicates the potential and effectiveness of the proposed model in comparison to the traditional single view clustering models, as well as other existing methods used in the literature for the studied datasets.

Subpathway Analysis based on Signaling-Pathway Impact Analysis of Signaling Pathway.

Pathway analysis is a common approach to gain insight from biological experiments. Signaling-pathway impact analysis (SPIA) is one such method and combines both the classical enrichment analysis and the actual perturbation on a given pathway. Because this method focuses on a single pathway, its resolution generally is not very high because the differentially expressed genes may be enriched in a local region of the pathway. In the present work, to identify cancer-related pathways, we incorporated a recent subpathway analysis method into the SPIA method to form the "sub-SPIA method." The original subpathway analysis uses the k-clique structure to define a subpathway. However, it is not sufficiently flexible to capture subpathways with complex structure and usually results in many overlapping subpathways. We therefore propose using the minimal-spanning-tree structure to find a subpathway. We apply this approach to colorectal cancer and lung cancer datasets, and our results show that sub-SPIA can identify many significant pathways associated with each specific cancer that other methods miss. Based on the entire pathway network in the Kyoto Encyclopedia of Genes and Genomes, we find that the pathways identified by sub-SPIA not only have the largest average degree, but also are more closely connected than those identified by other methods. This result suggests that the abnormality signal propagating through them might be responsible for the specific cancer or disease.

Gene expression profile in cardiovascular disease and preeclampsia: a meta-analysis of the transcriptome based on raw data from human studies deposited in Gene Expression Omnibus.

INTRODUCTION: Cardiovascular disease (CVD) and preeclampsia (PE) share common clinical features. We aimed to identify common transcriptomic signatures involved in CVD and PE in humans. METHODS: Meta-analysis of individual raw microarray data deposited in GEO, obtained from blood samples of patients with CVD versus controls and placental samples from women with PE versus healthy women with uncomplicated pregnancies. Annotation of cases versus control samples was taken directly from the microarray documentation. Genes that showed a significant differential expression in the majority of experiments were selected for subsequent analysis. Hypergeometric gene list analysis was performed using Bioconductor GOstats package. Bioinformatic analysis was performed in PANTHER. RESULTS: Seven studies in CVD and 5 studies in PE were eligible for meta-analysis. A total of 181 genes were found to be differentially expressed in microarray studies investigating gene expression in blood samples obtained from patients with CVD compared to controls and 925 genes were differentially expressed between preeclamptic and healthy placentas. Among these differentially expressed genes, 22 were common between CVD and PE. DISCUSSION: Bioinformatic analysis of these genes revealed oxidative stress, p-53 pathway feedback, inflammation mediated by chemokines and cytokines, interleukin signaling, B-cell activation, PDGF signaling, Wnt signaling, integrin signaling and Alzheimer disease pathways to be involved in the pathophysiology of both CVD and PE. Metabolism, development, response to stimulus, immune response and cell communication were the associated biologic processes in both conditions. Gene set enrichment analysis showed the following overlapping pathways between CVD and PE: TGF-ß-signaling, apoptosis, graft-versus-host disease, allograft rejection, chemokine signaling, steroid hormone synthesis, type I and II diabetes mellitus, VEGF signaling, pathways in cancer, GNRH signaling, Huntingtons disease and Notch signaling. CONCLUSION: CVD and PE share same common traits in their gene expression profile indicating common pathways in their pathophysiology.

Current Practice and Utility of Chromosome Microarray Analysis in Infants Undergoing Cardiac Surgery.

OBJECTIVE: Traditionally, karyotype and fluorescence in situ hybridization (FISH) were used for cytogenetic testing of infants with congenital heart disease (CHD) who underwent cardiac surgery at our institution. Recently, chromosome microarray analysis (CMA) has been performed in lieu of the traditional tests. A standardized approach to cytogenetic testing does not exist in this population. The purpose of this study was to assess the utility of CMA based on our current ordering practice. DESIGN: We reviewed the records of all infants (<1 year old) who underwent cardiac surgery at our institution from January 2010 to June 2013. Data included results of all cytogenetic testing performed. Diagnostic yield was calculated as the percentage of significant abnormal results obtained by each test modality. Patients were grouped by classification of CHD. RESULTS: Two hundred seventy-five (51%) of 535 infants who underwent cardiac surgery had cytogenetic testing. Of those tested, 154 (56%) had multiple tests performed and at least 18% were redundant or overlapping. The utilization of CMA has increased each year since its implementation. The diagnostic yield for karyotype, FISH and CMA was 10%, 12%, and 14%, respectively. CMA yield was significantly higher in patients with septal defects (33%, P = .01) compared with all other CHD classes. CMA detected abnormalities of unknown clinical significance in 13% of infants tested. CONCLUSIONS: In our center, redundant cytogenetic testing is frequently performed in infants undergoing cardiac surgery. The utilization of CMA has increased over time and abnormalities of unknown clinical significance are detected in an important subset of patients. A screening algorithm that risk-stratifies based on classification of CHD and clinical suspicion may provide a practical, data-driven approach to genetic testing in this population and limit unnecessary resource utilization.

Sample size considerations of prediction-validation methods in high-dimensional data for survival outcomes.

A variety of prediction methods are used to relate high-dimensional genome data with a clinical outcome using a prediction model. Once a prediction model is developed from a data set, it should be validated using a resampling method or an independent data set. Although the existing prediction methods have been intensively evaluated by many investigators, there has not been a comprehensive study investigating the performance of the validation methods, especially with a survival clinical outcome. Understanding the properties of the various validation methods can allow researchers to perform more powerful validations while controlling for type I error. In addition, sample size calculation strategy based on these validation methods is lacking. We conduct extensive simulations to examine the statistical properties of these validation strategies. In both simulations and a real data example, we have found that 10-fold cross-validation with permutation gave the best power while controlling type I error close to the nominal level. Based on this, we have also developed a sample size calculation method that will be used to design a validation study with a user-chosen combination of prediction. Microarray and genome-wide association studies data are used as illustrations. The power calculation method in this presentation can be used for the design of any biomedical studies involving high-dimensional data and survival outcomes.

[The emergence and development of gene expression profiling: a key component of the 3B (bench, bedside, bytes) in translational research]. / Les réseaux de l'expression génique - Émergence et développement d'un domaine clé de la génomique.

This paper examines the emergence and development of one of the key components of genomics, namely gene expression profiling. It does so by resorting to computer-based methods to analyze and visualize networks of scientific publications. Our results show the central role played by oncology in this domain, insofar as the initial proof-of-principle articles based on a plant model organism have quickly led to the demonstration of the value of these techniques in blood cancers and to applications in the field of solid tumors, and in particular breast cancer. The article also outlines the essential role played by novel bioinformatics and biostatistical tools in the development of the domain. These computational disciplines thus qualify as one of the three corners (in addition to the laboratory and the clinic) of the translational research triangle.

Customisation of the exome data analysis pipeline using a combinatorial approach.

The advent of next generation sequencing (NGS) technologies have revolutionised the way biologists produce, analyse and interpret data. Although NGS platforms provide a cost-effective way to discover genome-wide variants from a single experiment, variants discovered by NGS need follow up validation due to the high error rates associated with various sequencing chemistries. Recently, whole exome sequencing has been proposed as an affordable option compared to whole genome runs but it still requires follow up validation of all the novel exomic variants. Customarily, a consensus approach is used to overcome the systematic errors inherent to the sequencing technology, alignment and post alignment variant detection algorithms. However, the aforementioned approach warrants the use of multiple sequencing chemistry, multiple alignment tools, multiple variant callers which may not be viable in terms of time and money for individual investigators with limited informatics know-how. Biologists often lack the requisite training to deal with the huge amount of data produced by NGS runs and face difficulty in choosing from the list of freely available analytical tools for NGS data analysis. Hence, there is a need to customise the NGS data analysis pipeline to preferentially retain true variants by minimising the incidence of false positives and make the choice of right analytical tools easier. To this end, we have sampled different freely available tools used at the alignment and post alignment stage suggesting the use of the most suitable combination determined by a simple framework of pre-existing metrics to create significant datasets.

The tumor suppressor role of Src homology phosphotyrosine phosphatase 2 in hepatocellular carcinoma.

PURPOSE: The human gene PTPN11, which encodes the non-receptor protein tyrosine phosphatase of Src homology phosphotyrosine phosphatase 2 (Shp2), has been previously well interpreted as a proto-oncogene in a variety of malignancies. However, the tumor suppressor role of Shp2 has also been reported. The present study was conducted to investigate the role of Shp2 expression and its associated clinical manifestations in hepatocellular carcinoma (HCC). METHODS: A tissue microarray of 333 pairs of HCC and self-matched adjacent non-tumor tissues was constructed, and the expression of Shp2 was determined by immunohistochemistry. The results were also conformed by Western blotting and quantitative PCR of 31 self-paired fresh HCC specimens. The associations of Shp2 expression with 25 clinicopathologic features were analyzed. Overall survival analysis and multivariate analysis were performed. RESULTS: Significantly decreased Shp2 expression in tumor tissues (T) compared with adjacent non-tumor tissues (NT) could be detected, and the positive rate was 66.1 and 96.7%, respectively. We combined the T and NT Shp2 immunoreactivity by a variable of the decrease in Shp2 expression (ΔShp2) and divided cases into 2 groups: T < NT and T ≥ NT. Survival analysis showed both low Shp2 expression and T < NT group were significantly associated with short overall survival. Multivariate analysis showed ΔShp2 was an independent prognostic marker (P = 0.033; HR: 0.527; 95% CI: 0.293-0.950). CONCLUSION: Shp2 is a tumor suppressor, and the decrease in Shp2 expression was a new prognostic marker in HCC. The oncogenic role of Shp2 was tissue specific, and the therapeutic target of human gene PTPN11 should be reconsidered.

Classification of large microarray datasets using fast random forest construction.

Random forest is an ensemble classification algorithm. It performs well when most predictive variables are noisy and can be used when the number of variables is much larger than the number of observations. The use of bootstrap samples and restricted subsets of attributes makes it more powerful than simple ensembles of trees. The main advantage of a random forest classifier is its explanatory power: it measures variable importance or impact of each factor on a predicted class label. These characteristics make the algorithm ideal for microarray data. It was shown to build models with high accuracy when tested on high-dimensional microarray datasets. Current implementations of random forest in the machine learning and statistics community, however, limit its usability for mining over large datasets, as they require that the entire dataset remains permanently in memory. We propose a new framework, an optimized implementation of a random forest classifier, which addresses specific properties of microarray data, takes computational complexity of a decision tree algorithm into consideration, and shows excellent computing performance while preserving predictive accuracy. The implementation is based on reducing overlapping computations and eliminating dependency on the size of main memory. The implementation's excellent computational performance makes the algorithm useful for interactive data analyses and data mining.