Differential gene expression in disease: a comparison between high-throughput studies and the literature

Background

Differential gene expression is important to understand the biological differences between healthy and diseased states. Two common sources of differential gene expression data are microarray studies and the biomedical literature.

Methods

With the aid of text mining and gene expression analysis we have examined the comparative properties of these two sources of differential gene expression data.

Results

The literature shows a preference for reporting genes associated to higher fold changes in microarray data, rather than genes that are simply significantly differentially expressed. Thus, the resemblance between the literature and microarray data increases when the fold-change threshold for microarray data is increased. Moreover, the literature has a reporting preference for differentially expressed genes that (1) are overexpressed rather than underexpressed; (2) are overexpressed in multiple diseases; and (3) are popular in the biomedical literature at large. Additionally, the degree to which diseases are similar depends on whether microarray data or the literature is used to compare them. Finally, vaguely-qualified reports of differential expression magnitudes in the literature have only small correlation with microarray fold-change data.

Conclusions

Reporting biases of differential gene expression in the literature can be affecting our appreciation of disease biology and of the degree of similarity that actually exists between different diseases.

     

Multivariate search for differentially expressed gene combinations

Background

To identify differentially expressed genes, it is standard practice to test a two-sample hypothesis for each gene with a proper adjustment for multiple testing. Such tests are essentially univariate and disregard the multidimensional structure of microarray data. A more general two-sample hypothesis is formulated in terms of the joint distribution of any sub-vector of expression signals.

Results

By building on an earlier proposed multivariate test statistic, we propose a new algorithm for identifying differentially expressed gene combinations. The algorithm includes an improved random search procedure designed to generate candidate gene combinations of a given size. Cross-validation is used to provide replication stability of the search procedure. A permutation two-sample test is used for significance testing. We design a multiple testing procedure to control the family-wise error rate (FWER) when selecting significant combinations of genes that result from a successive selection procedure. A target set of genes is composed of all significant combinations selected via random search.

Conclusions

A new algorithm has been developed to identify differentially expressed gene combinations. The performance of the proposed search-and-testing procedure has been evaluated by computer simulations and analysis of replicated Affymetrix gene array data on age-related changes in gene expression in the inner ear of CBA mice.

     

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.

     

Time-course microarray analysis for identifying candidate genes involved in obesity-associated pathological changes in the mouse colon

Background

Obesity is known to increase the risk of colorectal cancer. However, mechanisms underlying the pathogenesis of obesity-induced colorectal cancer are not completely understood. The purposes of this study were to identify differentially expressed genes in the colon of mice with diet-induced obesity and to select candidate genes as early markers of obesity-associated abnormal cell growth in the colon.

Methods

C57BL/6N mice were fed normal diet (11% fat energy) or high-fat diet (40% fat energy) and were euthanized at different time points. Genome-wide expression profiles of the colon were determined at 2, 4, 8, and 12 weeks. Cluster analysis was performed using expression data of genes showing log₂ fold change of ≥1 or ≤?1 (twofold change), based on time-dependent expression patterns, followed by virtual network analysis.

Results

High-fat diet-fed mice showed significant increase in body weight and total visceral fat weight over 12 weeks. Time-course microarray analysis showed that 50, 47, 36, and 411 genes were differentially expressed at 2, 4, 8, and 12 weeks, respectively. Ten cluster profiles representing distinguishable patterns of genes differentially expressed over time were determined. Cluster 4, which consisted of genes showing the most significant alterations in expression in response to high-fat diet over 12 weeks, included Apoa4 (apolipoprotein A-IV), Ppap2b (phosphatidic acid phosphatase type 2B), Cel (carboxyl ester lipase), and Clps (colipase, pancreatic), which interacted strongly with surrounding genes associated with colorectal cancer or obesity.

Conclusions

Our data indicate that Apoa4, Ppap2b, Cel, and Clps are candidate early marker genes associated with obesity-related pathological changes in the colon. Genome-wide analyses performed in the present study provide new insights on selecting novel genes that may be associated with the development of diseases of the colon.

     

DTL reconciliation repair

Background

Maximum parsimony phylogenetic tree reconciliation is an important technique for reconstructing the evolutionary histories of hosts and parasites, genes and species, and other interdependent pairs. Since the problem of finding temporally feasible maximum parsimony reconciliations is NP-complete, current methods use either exact algorithms with exponential worst-case running time or heuristics that do not guarantee optimal solutions.

Results

We offer an efficient new approach that begins with a potentially infeasible maximum parsimony reconciliation and iteratively “repairs” it until it becomes temporally feasible.

Conclusions

In a non-trivial number of cases, this approach finds solutions that are better than those found by the widely-used Jane heuristic.

     

Microarray analysis reveals a potential role of LncRNAs expression in cardiac cell proliferation

Background

Long non-coding RNAs (LncRNAs) have been identified to play important roles in epigenetic processes that underpin organogenesis. However, the role of LncRNAs in the regulation of transition from fetal to adult life of human heart has not been evaluated.

Methods

Immunofiuorescent staining was used to determine the extent of cardiac cell proliferation. Human LncRNA microarrays were applied to define gene expression signatures of the fetal (13–17 weeks of gestation, n?=?4) and adult hearts (30–40 years old, n?=?4). Pathway analysis was performed to predict the function of differentially expressed mRNAs (DEM). DEM related to cell proliferation were selected to construct a lncRNA-mRNA co-expression network. Eight lncRNAs were confirmed by quantificational real-time polymerase chain reaction (n?=?6).

Results

Cardiac cell proliferation was significant in the fetal heart. Two thousand six hundred six lncRNAs and 3079 mRNAs were found to be differentially expressed. Cell cycle was the most enriched pathway in down-regulated genes in the adult heart. Eight lncRNAs (RP11-119 F7.5, AX747860, HBBP1, LINC00304, TPTE2P6, AC034193.5, XLOC_006934 and AL833346) were predicted to play a central role in cardiac cell proliferation.

Conclusions

We discovered a profile of lncRNAs differentially expressed between the human fetal and adult heart. Several meaningful lncRNAs involved in cardiac cell proliferation were disclosed.

     

Circular RNA expression profiles during the differentiation of mouse neural stem cells

Background

Circular RNAs (circRNAs) have recently been found to be expressed in human brain tissue, and many lines ofevidence indicate that circRNAs play regulatory roles in neurodevelopment. Proliferation and differentiation of neural stem cells (NSCs) are critical parts during development of central nervous system (CNS).To date, there have been no reports ofcircRNA expression profiles during the differentiation of mouse NSCs. We hypothesizethat circRNAs mayregulate gene expression in the proliferation anddifferentiation of NSCs.

Results

In this study, we obtained NSCs from the wild-type C57BL/6 J mouse fetal cerebral cortex. We extracted total RNA from NSCs in different differentiation stagesand then performed RNA-seq. By analyzing the RNA-Seq data, we found 37circRNAs and 4182 mRNAs differentially expressedduringthe NSC differentiation. Gene Ontology (GO) enrichment analysis of thecognate linear genes of these circRNAsrevealed that some enriched GO terms were related to neural activity. Furthermore, we performed a co-expression network analysis of these differentially expressed circRNAs and mRNAs. The result suggested a stronger GO enrichmentin neural features for both the cognate linear genes of circRNAs and differentially expressed mRNAs.

Conclusion

We performed the first circRNA investigation during the differentiation of mouse NSCs. Wefound that12 circRNAs might have regulatory roles duringthe NSC differentiation, indicating that circRNAs might be modulated during NSC differentiation.Our network analysis suggested the possible complex circRNA-mRNA mechanisms during differentiation, and future experimental workis need to validate these possible mechanisms.

     

Identification of biomarkers of chromophobe renal cell carcinoma by weighted gene co-expression network analysis

Background

Chromophobe renal cell carcinoma (ChRCC) is the second common subtype of non-clear cell renal cell carcinoma (nccRCC), which accounting for 4–5% of renal cell carcinoma (RCC). However, there is no effective bio-marker to predict clinical outcomes of this malignant disease. Bioinformatic methods may provide a feasible potential to solve this problem.

Methods

In this study, differentially expressed genes (DEGs) of ChRCC samples on The Cancer Genome Atlas database were filtered out to construct co-expression modules by weighted gene co-expression network analysis and the key module were identified by calculating module-trait correlations. Functional analysis was performed on the key module and candidate hub genes were screened out by co-expression and MCODE analysis. Afterwards, real hub genes were filter out in an independent dataset GSE15641 and validated by survival analysis.

Results

Overall 2215 DEGs were screened out to construct eight co-expression modules. Brown module was identified as the key module for the highest correlations with pathologic stage, neoplasm status and survival status. 29 candidate hub genes were identified. GO and KEGG analysis demonstrated most candidate genes were enriched in mitotic cell cycle. Three real hub genes (SKA1, ERCC6L, GTSE-1) were selected out after mapping candidate genes to GSE15641 and two of them (SKA1, ERCC6L) were significantly related to overall survivals of ChRCC patients.

Conclusions

In summary, our findings identified molecular markers correlated with progression and prognosis of ChRCC, which might provide new implications for improving risk evaluation, therapeutic intervention, and prognosis prediction in ChRCC patients.

     

A diVIsive Shuffling Approach (VIStA) for gene expression analysis to identify subtypes in Chronic Obstructive Pulmonary Disease

Background

An important step toward understanding the biological mechanisms underlying a complex disease is a refined understanding of its clinical heterogeneity. Relating clinical and molecular differences may allow us to define more specific subtypes of patients that respond differently to therapeutic interventions.

Results

We developed a novel unbiased method called diVIsive Shuffling Approach (VIStA) that identifies subgroups of patients by maximizing the difference in their gene expression patterns. We tested our algorithm on 140 subjects with Chronic Obstructive Pulmonary Disease (COPD) and found four distinct, biologically and clinically meaningful combinations of clinical characteristics that are associated with large gene expression differences. The dominant characteristic in these combinations was the severity of airflow limitation. Other frequently identified measures included emphysema, fibrinogen levels, phlegm, BMI and age. A pathway analysis of the differentially expressed genes in the identified subtypes suggests that VIStA is capable of capturing specific molecular signatures within in each group.

Conclusions

The introduced methodology allowed us to identify combinations of clinical characteristics that correspond to clear gene expression differences. The resulting subtypes for COPD contribute to a better understanding of its heterogeneity.

     

Evaluation of insulin-like growth factor-I in postmenopausal women with breast cancer treated with raloxifene

Background

The objective of this study was to evaluate serum IGF-I levels in postmenopausal women with breast cancer treated primarily with raloxifene.

Methods

Twenty-two postmenopausal patients with operable, stage I or II, estrogen receptor-positive carcinomas participated in this study. Following confirmation of diagnosis, the patients received 60 mg of raloxifene for 28 days prior to definitive surgery. Blood samples were collected for evaluation of serum IGF-I levels prior to initiating medication and following a 28-day treatment course. Student's t-test for paired samples was used in the statistical analysis. Significance was established at p < 0.05.

Results

Mean serum IGF-I levels pre- and post-raloxifene treatment were 143.7 ± 9.7 ng/ml and 94.8 ± 7.6 ng/ml, respectively. This reduction in serum IGF-I levels following treatment with raloxifene was statistically significant (p < 0.001).

Conclusion

Raloxifene significantly reduced serum IGF-I levels in postmenopausal women with breast cancer.

     

Identification of novel biomarker candidates by proteomic analysis of cerebrospinal fluid from patients with moyamoya disease using SELDI-TOF-MS

Background

Moyamoya disease (MMD) is an uncommon cerebrovascular condition with unknown etiology characterized by slowly progressive stenosis or occlusion of the bilateral internal carotid arteries associated with an abnormal vascular network. MMD is a major cause of stroke, specifically in the younger population. Diagnosis is based on only radiological features as no other clinical data are available. The purpose of this study was to identify novel biomarker candidate proteins differentially expressed in the cerebrospinal fluid (CSF) of patients with MMD using proteomic analysis.

Methods

For detection of biomarkers, CSF samples were obtained from 20 patients with MMD and 12 control patients. Mass spectral data were generated by surface-enhanced laser desorption/ionization time-of-flight mass spectrometry (SELDI-TOF-MS) with an anion exchange chip in three different buffer conditions. After expression difference mapping was undertaken using the obtained protein profiles, a comparative analysis was performed.

Results

A statistically significant number of proteins (34) were recognized as single biomarker candidate proteins which were differentially detected in the CSF of patients with MMD, compared to the control patients (p < 0.05). All peak intensity profiles of the biomarker candidates underwent classification and regression tree (CART) analysis to produce prediction models. Two important biomarkers could successfully classify the patients with MMD and control patients.

Conclusions

In this study, several novel biomarker candidate proteins differentially expressed in the CSF of patients with MMD were identified by a recently developed proteomic approach. This is a pilot study of CSF proteomics for MMD using SELDI technology. These biomarker candidates have the potential to shed light on the underlying pathogenesis of MMD.

     

PROPER: Performance visualization for optimizing and comparing ranking classifiers in MATLAB

Background

One of the recent challenges of computational biology is development of new algorithms, tools and software to facilitate predictive modeling of big data generated by high-throughput technologies in biomedical research.

Results

To meet these demands we developed PROPER - a package for visual evaluation of ranking classifiers for biological big data mining studies in the MATLAB environment.

Conclusion

PROPER is an efficient tool for optimization and comparison of ranking classifiers, providing over 20 different two- and three-dimensional performance curves.