An efficient and robust statistical modeling approach to discover differentially expressed genes using genomic expression profiles

T-type voltage-gated channels, Na+/Ca2+-exchanger, and calpain-2 promote photoreceptor cell death in inherited retinal degeneration

Inherited retinal degenerations (IRDs) are a group of untreatable and commonly blinding diseases characterized by progressive photoreceptor loss. IRD pathology has been linked to an excessive activation of cyclic nucleotide-gated channels (CNGC) leading to Na+- and Ca2+-influx, subsequent activation of voltage-gated Ca2+-channels (VGCC), and further Ca2+ influx. However, a connection between excessive Ca2+ influx and photoreceptor loss has yet to be proven.Here, we used whole-retina and single-cell RNA-sequencing to compare gene expression between the rd1 mouse model for IRD and wild-type (wt) mice. Differentially expressed genes indicated links to several Ca2+-signalling related pathways. To explore these, rd1 and wt organotypic retinal explant cultures were treated with the intracellular Ca2+-chelator BAPTA-AM or inhibitors of different Ca2+-permeable channels, including CNGC, L-type VGCC, T-type VGCC, Ca2+-release-activated channel (CRAC), and Na+/Ca2+ exchanger (NCX). Moreover, we employed the novel compound NA-184 to selectively inhibit the Ca2+-dependent protease calpain-2. Effects on the retinal activity of poly(ADP-ribose) polymerase (PARP), sirtuin-type histone-deacetylase, calpains, as well as on activation of calpain-1, and - 2 were monitored, cell death was assessed via the TUNEL assay.While rd1 photoreceptor cell death was reduced by BAPTA-AM, Ca2+-channel blockers had divergent effects: While inhibition of T-type VGCC and NCX promoted survival, blocking CNGCs and CRACs did not. The treatment-related activity patterns of calpains and PARPs corresponded to the extent of cell death. Remarkably, sirtuin activity and calpain-1 activation were linked to photoreceptor protection, while calpain-2 activity was related to degeneration. In support of this finding, the calpain-2 inhibitor NA-184 protected rd1 photoreceptors.These results suggest that Ca2+ overload in rd1 photoreceptors may be triggered by T-type VGCCs and NCX. High Ca2+-levels likely suppress protective activity of calpain-1 and promote retinal degeneration via activation of calpain-2. Overall, our study details the complexity of Ca2+-signalling in photoreceptors and emphasizes the importance of targeting degenerative processes specifically to achieve a therapeutic benefit for IRDs. Video Abstract.

Genetic and histopathological analysis of spermatogenesis after short-term testicular torsion in rats

BACKGROUND

Patients with testicular torsion (TT) may exhibit impaired spermatogenesis from reperfusion injury after detorsion surgery. Alteration in the expressions of spermatogenesis-related genes induced by TT have not been fully elucidated.

METHODS

Eight-week-old Sprague-Dawley rats were grouped as follows: group 1 (sham-operated), group 2 (TT without reperfusion) and group 3 (TT with reperfusion). TT was induced by rotating the left testis 720° for 1 h. Testicular reperfusion proceeded for 24 h. Histopathological examination, oxidative stress biomarker measurements, RNA sequencing and RT-PCR were performed.

RESULTS

Testicular ischemia/reperfusion injury induced marked histopathological changes. Germ cell apoptosis was significantly increased in group 3 compared with group 1 and 2 (mean apoptotic index: 26.22 vs. 0.64 and 0.56; p = 0.024, and p = 0.024, respectively). Johnsen score in group 3 was smaller than that in group 1 and 2 (mean: 8.81 vs 9.45 and 9.47 points/tubule; p = 0.001, p < 0.001, respectively). Testicular ischemia/reperfusion injury significantly upregulated the expression of genes associated with apoptosis and antioxidant enzymes and significantly downregulated the expression of genes associated with spermatogenesis.

CONCLUSION

One hour of TT followed by reperfusion injury caused histopathological testicular damage. The relatively high Johnsen score indicated spermatogenesis was maintained. Genes associated with spermatogenesis were downregulated in the TT rat model.

IMPACT

How ischemia/reperfusion injury in testicular torsion (TT) affects the expressions of genes associated with spermatogenesis has not been fully elucidated. This is the first study to report comprehensive gene expression profiles using next generation sequencing for an animal model of TT. Our results revealed that ischemia/reperfusion injury downregulated the expression of genes associated with spermatogenesis and sperm function in addition to histopathological damage, even though the duration of ischemia was short.

Patterns of differential expression by association in omic data using a new measure based on ensemble learning

The ongoing development of high-throughput technologies is allowing the simultaneous monitoring of the expression levels for hundreds or thousands of biological inputs with the proliferation of what has been coined as omic data sources. One relevant issue when analyzing such data sources is concerned with the detection of differential expression across two experimental conditions, clinical status or two classes of a biological outcome. While a great deal of univariate data analysis approaches have been developed to address the issue, strategies for assessing interaction patterns of differential expression are scarce in the literature and have been limited to ad hoc solutions. This paper contributes to the problem by exploiting the facilities of an ensemble learning algorithm like random forests to propose a measure that assesses the differential expression explained by the interaction of the omic variables so subtle biological patterns may be uncovered as a result. The out of bag error rate, which is an estimate of the predictive accuracy of a random forests classifier, is used as a by-product to propose a new measure that assesses interaction patterns of differential expression. Its performance is studied in synthetic scenarios and it is also applied to real studies on SARS-CoV-2 and colon cancer data where it uncovers associations that remain undetected by other methods. Our proposal is aimed at providing a novel approach that may help the experts in biomedical and life sciences to unravel insightful interaction patterns that may decipher the molecular mechanisms underlying biological and clinical outcomes.

Flammulina filiformis Pkac Gene Complementing in Neurospora crassa Mutant Reveals Its Function in Mycelial Growth and Abiotic Stress Response

Flammulina filiformis is a popular edible mushroom that easily suffers from heat and oxidative stresses. The cyclic adenylate-dependent protein kinase A (cAMP/PKA) pathway is the main signaling pathway in response to environmental stress, and the PKAC is the terminal catalytic subunit of this pathway. In this study, the Pkac gene was identified in F. filiformis, which was highly conserved in basidiomycetes and ascomycetes. The transcription analysis showed that the Pkac gene was involved in the mycelial growth and the fruiting body development of fungi. In Neurospora crassa, the Pkac gene deletion (ΔPkac) resulted in the slower growth of the mycelia. We complemented the F. filiformis FfPkac to N. crassa ΔPkac mutant to obtain the CPkac strain. The mycelial growth in the CPkac strain was restored to the same level as the WT strain. In addition, the FfPkac gene showed significantly up-regulated expression under heat and oxidative stresses. By analyzing the differentially expressed genes of ΔPkac and Cpkac with WT, respectively, seven downstream genes regulated by Pkac were identified and may be related to mycelial growth. They were mainly focused on microbial metabolism in diverse environments, mitochondrial biogenesis, protein translation and nucleocytoplasmic transport. RT-qPCR results confirmed that the expression patterns of these seven genes were consistent with FfPkac under heat and oxidative stresses. The results revealed the conserved functions of PKAC in filamentous fungi and its regulatory mechanism in response to heat and oxidative stresses.

Computationally repurposing drugs for breast cancer subtypes using a network-based approach

‘De novo’ drug discovery is costly, slow, and with high risk. Repurposing known drugs for treatment of other diseases offers a fast, low-cost/risk and highly-efficient method toward development of efficacious treatments. The emergence of large-scale heterogeneous biomolecular networks, molecular, chemical and bioactivity data, and genomic and phenotypic data of pharmacological compounds is enabling the development of new area of drug repurposing called ‘in silico’ drug repurposing, i.e., computational drug repurposing (CDR). The aim of CDR is to discover new indications for an existing drug (drug-centric) or to identify effective drugs for a disease (disease-centric). Both drug-centric and disease-centric approaches have the common challenge of either assessing the similarity or connections between drugs and diseases. However, traditional CDR is fraught with many challenges due to the underlying complex pharmacology and biology of diseases, genes, and drugs, as well as the complexity of their associations. As such, capturing highly non-linear associations among drugs, genes, diseases by most existing CDR methods has been challenging. We propose a network-based integration approach that can best capture knowledge (and complex relationships) contained within and between drugs, genes and disease data. A network-based machine learning approach is applied thereafter by using the extracted knowledge and relationships in order to identify single and pair of approved or experimental drugs with potential therapeutic effects on different breast cancer subtypes. Indeed, further clinical analysis is needed to confirm the therapeutic effects of identified drugs on each breast cancer subtype.

Systems Biology and Bioinformatics approach to Identify blood based signatures molecules and drug targets of patient with COVID-19

Severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) infection results in the development of a highly contagious respiratory ailment known as new coronavirus disease (COVID-19). Despite the fact that the prevalence of COVID-19 continues to rise, it is still unclear how people become infected with SARS-CoV-2 and how patients with COVID-19 become so unwell. Detecting biomarkers for COVID-19 using peripheral blood mononuclear cells (PBMCs) may aid in drug development and treatment. This research aimed to find blood cell transcripts that represent levels of gene expression associated with COVID-19 progression. Through the development of a bioinformatics pipeline, two RNA-Seq transcriptomic datasets and one microarray dataset were studied and discovered 102 significant differentially expressed genes (DEGs) that were shared by three datasets derived from PBMCs. To identify the roles of these DEGs, we discovered disease-gene association networks and signaling pathways, as well as we performed gene ontology (GO) studies and identified hub protein. Identified significant gene ontology and molecular pathways improved our understanding of the pathophysiology of COVID-19, and our identified blood-based hub proteins TPX2, DLGAP5, NCAPG, CCNB1, KIF11, HJURP, AURKB, BUB1B, TTK, and TOP2A could be used for the development of therapeutic intervention. In COVID-19 subjects, we discovered effective putative connections between pathological processes in the transcripts blood cells, suggesting that blood cells could be used to diagnose and monitor the disease’s initiation and progression as well as developing drug therapeutics.

Osteoblasts Promote Prostate Cancer Cell Proliferation Through Androgen Receptor Independent Mechanisms

Patients with metastatic prostate cancer frequently develop bone metastases that elicit significant skeletal morbidity and increased mortality. The high tropism of prostate cancer cells for bone and their tendency to induce the osteoblastic-like phenotype are a result of a complex interplay between tumor cells and osteoblasts. Although the role of osteoblasts in supporting prostate cancer cell proliferation has been reported by previous studies, their precise contribution in tumor growth remains to be fully elucidated. Here, we tried to dissect the molecular signaling underlining the interactions between castration-resistant prostate cancer (CRPC) cells and osteoblasts using in vitro co-culture models. Transcriptomic analysis showed that osteoblast-conditioned media (OCM) induced the overexpression of genes related to cell cycle in the CRPC cell line C4-2B but, surprisingly, reduced androgen receptor (AR) transcript levels. In-depth analysis of AR expression in C4-2B cells after OCM treatment showed an AR reduction at the mRNA (p = 0.0047), protein (p = 0.0247), and functional level (p = 0.0029) and, concomitantly, an increase of C4-2B cells in S-G2-M cell cycle phases (p = 0.0185). An extensive proteomic analysis revealed in OCM the presence of some molecules that reduced AR activation, and among these, Matrix metalloproteinase-1 (MMP-1) was the only one able to block AR function (0.1 ng/ml p = 0.006; 1 ng/ml p = 0.002; 10 ng/ml p = 0.0001) and, at the same time, enhance CRPC proliferation (1 ng/ml p = 0.009; 10 ng/ml p = 0.033). Although the increase of C4-2B cell growth induced by MMP-1 did not reach the proliferation levels observed after OCM treatment, the addition of Vorapaxar, an MMP-1 receptor inhibitor (Protease-activated receptor-1, PAR-1), significantly reduced C4-2B cell cycle (0.1 μM p = 0.014; 1 μM p = 0.0087). Overall, our results provide a novel AR-independent mechanism of CRPC proliferation and suggest that MMP-1/PAR-1 could be one of the potential pathways involved in this process.

New Radiomic Markers of Pulmonary Vein Morphology Associated With Post-Ablation Recurrence of Atrial Fibrillation

Objective: To identify radiomic and clinical features associated with post-ablation recurrence of AF, given that cardiac morphologic changes are associated with persistent atrial fibrillation (AF), and initiating triggers of AF often arise from the pulmonary veins which are targeted in ablation. Methods: Subjects with pre-ablation contrast CT scans prior to first-time catheter ablation for AF between 2014-2016 were retrospectively identified. A training dataset (D1) was constructed from left atrial and pulmonary vein morphometric features extracted from equal numbers of consecutively included subjects with and without AF recurrence determined at 1 year. The top-performing combination of feature selection and classifier methods based on C-statistic was evaluated on a validation dataset (D2), composed of subjects retrospectively identified between 2005-2010. Clinical models ([Formula: see text]) were similarly evaluated and compared to radiomic ([Formula: see text]) and radiomic-clinical models ([Formula: see text]), each independently validated on D2. Results: Of 150 subjects in D1, 108 received radiofrequency ablation and 42 received cryoballoon. Radiomic features of recurrence included greater right carina angle, reduced anterior-posterior atrial diameter, greater atrial volume normalized to height, and steeper right inferior pulmonary vein angle. Clinical features predicting recurrence included older age, greater BMI, hypertension, and warfarin use; apixaban use was associated with reduced recurrence. AF recurrence was predicted with radio-frequency ablation models on D2 subjects with C-statistics of 0.68, 0.63, and 0.70 for radiomic, clinical, and combined feature models, though these were not prognostic in patients treated with cryoballoon. Conclusions: Pulmonary vein morphology associated with increased likelihood of AF recurrence within 1 year of catheter ablation was identified on cardiac CT. Significance: Radiomic and clinical features-based predictive models may assist in identifying atrial fibrillation ablation candidates with greatest likelihood of successful outcome.

Gene Correlation Guided Gene Selection for Microarray Data Classification

The microarray cancer data obtained by DNA microarray technology play an important role for cancer prevention, diagnosis, and treatment. However, predicting the different types of tumors is a challenging task since the sample size in microarray data is often small but the dimensionality is very high. Gene selection, which is an effective means, is aimed at mitigating the curse of dimensionality problem and can boost the classification accuracy of microarray data. However, many of previous gene selection methods focus on model design, but neglect the correlation between different genes. In this paper, we introduce a novel unsupervised gene selection method by taking the gene correlation into consideration, named gene correlation guided gene selection (G³CS). Specifically, we calculate the covariance of different gene dimension pairs and embed it into our unsupervised gene selection model to regularize the gene selection coefficient matrix. In such a manner, redundant genes can be effectively excluded. In addition, we utilize a matrix factorization term to exploit the cluster structure of original microarray data to assist the learning process. We design an iterative updating algorithm with convergence guarantee to solve the resultant optimization problem. Experimental results on six publicly available microarray datasets are conducted to validate the efficacy of our proposed method.

Network-based multi-task learning models for biomarker selection and cancer outcome prediction

MOTIVATION

Detecting cancer gene expression and transcriptome changes with mRNA-sequencing or array-based data are important for understanding the molecular mechanisms underlying carcinogenesis and cellular events during cancer progression. In previous studies, the differentially expressed genes were detected across patients in one cancer type. These studies ignored the role of mRNA expression changes in driving tumorigenic mechanisms that are either universal or specific in different tumor types. To address the problem, we introduce two network-based multi-task learning frameworks, NetML and NetSML, to discover common differentially expressed genes shared across different cancer types as well as differentially expressed genes specific to each cancer type. The proposed frameworks consider the common latent gene co-expression modules and gene-sample biclusters underlying the multiple cancer datasets to learn the knowledge crossing different tumor types.

RESULTS

Large-scale experiments on simulations and real cancer high-throughput datasets validate that the proposed network-based multi-task learning frameworks perform better sample classification compared with the models without the knowledge sharing across different cancer types. The common and cancer-specific molecular signatures detected by multi-task learning frameworks on The Cancer Genome Atlas ovarian, breast and prostate cancer datasets are correlated with the known marker genes and enriched in cancer-relevant Kyoto Encyclopedia of Genes and Genome pathways and gene ontology terms.

AVAILABILITY AND IMPLEMENTATION

Source code is available at: https://github.com/compbiolabucf/NetML.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Single-Cell Classification Using Mass Spectrometry through Interpretable Machine Learning

The brain consists of organized ensembles of cells that exhibit distinct morphologies, cellular connectivity, and dynamic biochemistries that control the executive functions of an organism. However, the relationships between chemical heterogeneity, cell function, and phenotype are not always understood. Recent advancements in matrix-assisted laser desorption/ionization mass spectrometry have enabled the high-throughput, multiplexed chemical analysis of single cells, capable of resolving hundreds of molecules in each mass spectrum. We developed a machine learning workflow to classify single cells according to their mass spectra based on cell groups of interest (GOI), e.g., neurons vs astrocytes. Three data sets from various cell groups were acquired on three different mass spectrometer platforms representing thousands of individual cell spectra that were collected and used to validate the single cell classification workflow. The trained models achieved >80% classification accuracy and were subjected to the recently developed instance-based model interpretation framework, SHapley Additive exPlanations (SHAP), which locally assigns feature importance for each single-cell spectrum. SHAP values were used for both local and global interpretations of our data sets, preserving the chemical heterogeneity uncovered by the single-cell analysis while offering the ability to perform supervised analysis. The top contributing mass features to each of the GOI were ranked and selected using mean absolute SHAP values, highlighting the features that are specific to the defined GOI. Our approach provides insight into discriminating the chemical profiles of the single cells through interpretable machine learning, facilitating downstream analysis and validation.

Machine learning technology in the application of genome analysis: A systematic review

Machine learning (ML) is a powerful technique to tackle many problems in data mining and predictive analytics. We believe that ML will be of considerable potentials in the field of bioinformatics since the high-throughput technology is producing ever increasing biological data. In this review, we summarized major ML algorithms and conditions that must be paid attention to when applying these algorithms to genomic problems in details and we provided a list of examples from different perspectives and data analysis challenges at present.

Three-way interaction model with switching mechanism as an effective strategy for tracing functionally-related genes

Introduction: Identification of functionally-related genes is an important step in understanding biological systems. The most popular strategy to infer functional dependence is to study pairwise correlations between gene expression levels. However, certain functionally-related genes may have a low expression correlation due to their nonlinear interactions. The use of a three-way interaction (3WI) model with switching mechanism (SM) is a relatively new strategy to trace functionally-related genes. The 3WI model traces the dynamic and nonlinear nature of the co-expression relationship of two genes by introducing their link to the expression level of a third gene. Areas covered: In this paper, we reviewed a variety of existing methods for tracing the 3WIs. Furthermore, we provide a comprehensive review of the previous biological studies based on 3WI models. Expert commentary: Comparison of features of these methods indicates that the modified liquid association algorithm has the best efficiency for tracing 3WI between others. The limited number of biological studies based on the 3WI suggests that high computational demand of the available algorithms is a major challenge to apply this approach for analyzing high-throughput omics data.

Analysis of high-throughput biological data using their rank values

High-throughput biological technologies are routinely used to generate gene expression profiling or cytogenetics data. To achieve high performance, methods available in the literature become more specialized and often require high computational resources. Here, we propose a new versatile method based on the data-ordering rank values. We use linear algebra, the Perron-Frobenius theorem and also extend a method presented earlier for searching differentially expressed genes for the detection of recurrent copy number aberration. A result derived from the proposed method is a one-sample Student's t-test based on rank values. The proposed method is to our knowledge the only that applies to gene expression profiling and to cytogenetics data sets. This new method is fast, deterministic, and requires a low computational load. Probabilities are associated with genes to allow a statistically significant subset selection in the data set. Stability scores are also introduced as quality parameters. The performance and comparative analyses were carried out using real data sets. The proposed method can be accessed through an R package available from the CRAN (Comprehensive R Archive Network) website: https://cran.r-project.org/web/packages/fcros .

Investigating a multigene prognostic assay based on significant pathways for Luminal A breast cancer through gene expression profile analysis

The present study aimed to investigate potential recurrence-risk biomarkers based on significant pathways for Luminal A breast cancer through gene expression profile analysis. Initially, the gene expression profiles of Luminal A breast cancer patients were downloaded from The Cancer Genome Atlas database. The differentially expressed genes (DEGs) were identified using a Limma package and the hierarchical clustering analysis was conducted for the DEGs. In addition, the functional pathways were screened using Kyoto Encyclopedia of Genes and Genomes pathway enrichment analyses and rank ratio calculation. The multigene prognostic assay was exploited based on the statistically significant pathways and its prognostic function was tested using train set and verified using the gene expression data and survival data of Luminal A breast cancer patients downloaded from the Gene Expression Omnibus. A total of 300 DEGs were identified between good and poor outcome groups, including 176 upregulated genes and 124 downregulated genes. The DEGs may be used to effectively distinguish Luminal A samples with different prognoses verified by hierarchical clustering analysis. There were 9 pathways screened as significant pathways and a total of 18 DEGs involved in these 9 pathways were identified as prognostic biomarkers. According to the survival analysis and receiver operating characteristic curve, the obtained 18-gene prognostic assay exhibited good prognostic function with high sensitivity and specificity to both the train and test samples. In conclusion the 18-gene prognostic assay including the key genes, transcription factor 7-like 2, anterior parietal cortex and lymphocyte enhancer factor-1 may provide a new method for predicting outcomes and may be conducive to the promotion of precision medicine for Luminal A breast cancer.

Applications of Support Vector Machine (SVM) Learning in Cancer Genomics

Machine learning with maximization (support) of separating margin (vector), called support vector machine (SVM) learning, is a powerful classification tool that has been used for cancer genomic classification or subtyping. Today, as advancements in high-throughput technologies lead to production of large amounts of genomic and epigenomic data, the classification feature of SVMs is expanding its use in cancer genomics, leading to the discovery of new biomarkers, new drug targets, and a better understanding of cancer driver genes. Herein we reviewed the recent progress of SVMs in cancer genomic studies. We intend to comprehend the strength of the SVM learning and its future perspective in cancer genomic applications.

Gene selection for microarray data classification via subspace learning and manifold regularization

With the rapid development of DNA microarray technology, large amount of genomic data has been generated. Classification of these microarray data is a challenge task since gene expression data are often with thousands of genes but a small number of samples. In this paper, an effective gene selection method is proposed to select the best subset of genes for microarray data with the irrelevant and redundant genes removed. Compared with original data, the selected gene subset can benefit the classification task. We formulate the gene selection task as a manifold regularized subspace learning problem. In detail, a projection matrix is used to project the original high dimensional microarray data into a lower dimensional subspace, with the constraint that the original genes can be well represented by the selected genes. Meanwhile, the local manifold structure of original data is preserved by a Laplacian graph regularization term on the low-dimensional data space. The projection matrix can serve as an importance indicator of different genes. An iterative update algorithm is developed for solving the problem. Experimental results on six publicly available microarray datasets and one clinical dataset demonstrate that the proposed method performs better when compared with other state-of-the-art methods in terms of microarray data classification. Graphical Abstract The graphical abstract of this work.

An improved analysis methodology for translational profiling by microarray

Translational regulation plays a central role in the global gene expression of a cell, and detection of such regulation has allowed deciphering of critical biological mechanisms. Genome-wide studies of the regulation of translation (translatome) performed on microarrays represent a substantial proportion of studies, alongside with recent advances in deep-sequencing methods. However, there has been a lack of development in specific processing methodologies that deal with the distinct nature of translatome array data. In this study, we confirm that polysome profiling yields skewed data and thus violates the conventional transcriptome analysis assumptions. Using a comprehensive simulation of translatome array data varying the percentage and symmetry of deregulation, we show that conventional analysis methods (Quantile and LOESS normalizations) and statistical tests failed, respectively, to correctly normalize the data and to identify correctly deregulated genes (DEGs). We thus propose a novel analysis methodology available as a CRAN package; Internal Control Analysis of Translatome (INCATome) based on a normalization tied to a group of invariant controls. We confirm that INCATome outperforms the other normalization methods and allows a stringent identification of DEGs. More importantly, INCATome implementation on a biological translatome data set (cells silenced for splicing factor PSF) resulted in the best normalization performance and an improved validation concordance for identification of true positive DEGs. Finally, we provide evidence that INCATome is able to infer novel biological pathways with superior discovery potential, thus confirming the benefits for researchers of implementing INCATome for future translatome studies as well as for existing data sets to generate novel avenues for research.

How many differentially expressed genes: A perspective from the comparison of genotypic and phenotypic distances

Identifying differentially expressed genes is critical in microarray data analysis. Many methods have been developed by combining p-value, fold-change, and various statistical models to determine these genes. When using these methods, it is necessary to set up various pre-determined cutoff values. However, many of these cutoff values are somewhat arbitrary and may not have clear connections to biology. In this study, a genetic distance method based on gene expression level was developed to analyze eight sets of microarray data extracted from the GEO database. Since the genes used in distance calculation have been ranked by fold-change, the genetic distance becomes more stable when adding more genes in the calculation, indicating there is an optimal set of genes which are sufficient to characterize the stable difference between samples. This set of genes is differentially expressed genes representing both the genotypic and phenotypic differences between samples.

The impact of RNA sequence library construction protocols on transcriptomic profiling of leukemia

Background

RNA sequencing (RNA-seq) has become an indispensable tool to identify disease associated transcriptional profiles and determine the molecular underpinnings of diseases. However, the broad adaptation of the methodology into the clinic is still hampered by inconsistent results from different RNA-seq protocols and involves further evaluation of its analytical reliability using patient samples. Here, we applied two commonly used RNA-seq library preparation protocols to samples from acute leukemia patients to understand how poly-A-tailed mRNA selection (PA) and ribo-depletion (RD) based RNA-seq library preparation protocols affect gene fusion detection, variant calling, and gene expression profiling.

Results

Overall, the protocols produced similar results with consistent outcomes. Nevertheless, the PA protocol was more efficient in quantifying expression of leukemia marker genes and showed better performance in the expression-based classification of leukemia. Independent qRT-PCR experiments verified that the PA protocol better represented total RNA compared to the RD protocol. In contrast, the RD protocol detected a higher number of non-coding RNA features and had better alignment efficiency. The RD protocol also recovered more known fusion-gene events, although variability was seen in fusion gene predictions.

Conclusion

The overall findings provide a framework for the use of RNA-seq in a precision medicine setting with limited number of samples and suggest that selection of the library preparation protocol should be based on the objectives of the analysis.

Electronic supplementary material

The online version of this article (doi:10.1186/s12864-017-4039-1) contains supplementary material, which is available to authorized users.

Pathway Interaction Network Analysis Identifies Dysregulated Pathways in Human Monocytes Infected by Listeria monocytogenes

In our study, we aimed to extract dysregulated pathways in human monocytes infected by Listeria monocytogenes (LM) based on pathway interaction network (PIN) which presented the functional dependency between pathways. After genes were aligned to the pathways, principal component analysis (PCA) was used to calculate the pathway activity for each pathway, followed by detecting seed pathway. A PIN was constructed based on gene expression profile, protein-protein interactions (PPIs), and cellular pathways. Identifying dysregulated pathways from the PIN was performed relying on seed pathway and classification accuracy. To evaluate whether the PIN method was feasible or not, we compared the introduced method with standard network centrality measures. The pathway of RNA polymerase II pretranscription events was selected as the seed pathway. Taking this seed pathway as start, one pathway set (9 dysregulated pathways) with AUC score of 1.00 was identified. Among the 5 hub pathways obtained using standard network centrality measures, 4 pathways were the common ones between the two methods. RNA polymerase II transcription and DNA replication owned a higher number of pathway genes and DEGs. These dysregulated pathways work together to influence the progression of LM infection, and they will be available as biomarkers to diagnose LM infection.

A clustering-based approach for efficient identification of microRNA combinatorial biomarkers

BACKGROUND

MicroRNAs (miRNAs) have great potential serving as tumor biomarkers and therapeutic targets. As the rapid development of high-throughput experimental technology, gene expression experiments have become more and more specialized and diversified. The complex data structure has brought great challenge for the identification of biomarkers. In the meantime, current statistical and machine learning methods for detecting biomarkers have the problem of low reliability and biased criteria.

RESULTS

This study aims to select combinatorial miRNA biomarkers, which have higher sensitivity and specificity than single-gene biomarkers. In order to avoid exhaustive search and redundant information, miRNAs are firstly clustered, then the combinations of representative cluster members are assessed as potential biomarkers. Both the criteria for the partition of clusters and selection of representative members are based on Fisher linear discriminant analysis (FDA). The FDA-based criterion has been demonstrated to be superior to three other criteria in selecting representative members, and also good at refining clusters. In the comparison with eight common feature selection methods, this clustering-based method performs the best with regard to the discriminative ability of selected biomarkers.

CONCLUSIONS

Our experimental results demonstrate that the clustering-based method can identify microRNA combinatorial biomarkers with high accuracy and efficiency. Our method and data are available to the public upon request.

SMITE: an R/Bioconductor package that identifies network modules by integrating genomic and epigenomic information

BACKGROUND

The molecular assays that test gene expression, transcriptional, and epigenetic regulation are increasingly diverse and numerous. The information generated by each type of assay individually gives an insight into the state of the cells tested. What should be possible is to add the information derived from separate, complementary assays to gain higher-confidence insights into cellular states. At present, the analysis of multi-dimensional, massive genome-wide data requires an initial pruning step to create manageable subsets of observations that are then used for integration, which decreases the sizes of the intersecting data sets and the potential for biological insights. Our Significance-based Modules Integrating the Transcriptome and Epigenome (SMITE) approach was developed to integrate transcriptional and epigenetic regulatory data without a loss of resolution.

RESULTS

SMITE combines p-values by accounting for the correlation between non-independent values within data sets, allowing genes and gene modules in an interaction network to be assigned significance values. The contribution of each type of genomic data can be weighted, permitting integration of individually under-powered data sets, increasing the overall ability to detect effects within modules of genes. We apply SMITE to a complex genomic data set including the epigenomic and transcriptomic effects of Toxoplasma gondii infection on human host cells and demonstrate that SMITE is able to identify novel subnetworks of dysregulated genes. Additionally, we show that SMITE outperforms Functional Epigenetic Modules (FEM), the current paradigm of using the spin-glass algorithm to integrate gene expression and epigenetic data.

CONCLUSIONS

SMITE represents a flexible, scalable tool that allows integration of transcriptional and epigenetic regulatory data from genome-wide assays to boost confidence in finding gene modules reflecting altered cellular states.

Identification of Novel Signal Transduction, Immune Function, and Oxidative Stress Genes and Pathways by Topiramate for Treatment of Methamphetamine Dependence Based on Secondary Outcomes

BACKGROUND

Topiramate (TPM) is suggested to be a promising medication for treatment of methamphetamine (METH) dependence, but the molecular basis remains to be elucidated.

METHODS

Among 140 METH-dependent participants randomly assigned to receive either TPM (N = 69) or placebo (N = 71) in a previously conducted randomized controlled trial, 50 TPM- and 49 placebo-treated participants had a total 212 RNA samples available at baseline, week 8, and week 12 time points. Following our primary analysis of gene expression data, we reanalyzed the microarray expression data based on a latent class analysis of binary secondary outcomes during weeks 1-12 that provided a classification of 21 responders and 31 non-responders with consistent responses at both time points.

RESULTS

Based on secondary outcomes, 1,381, 576, 905, and 711 differentially expressed genes at nominal P values < 0.05 were identified in responders versus non-responders for week 8 TPM, week 8 placebo, week 12 TPM, and week 12 placebo groups, respectively. Among 1,381 genes identified in week 8 TPM responders, 359 genes were identified in both week 8 and week 12 TPM groups, of which 300 genes were exclusively detected in TPM responders. Of them, 32 genes had nominal P values < 5 × 10^-3 at either week 8 or week 12 and false discovery rates < 0.15 at both time points with consistent directions of gene expression changes, which include GABARAPL1, GPR155, and IL15RA in GABA receptor signaling that represent direct targets for TPM. Analyses of these 300 genes revealed 7 enriched pathways belonging to neuronal function/synaptic plasticity, signal transduction, inflammation/immune function, and oxidative stress response categories. No pathways were enriched for 72 genes exclusively detected in both week 8 and week 12 placebo groups.

CONCLUSION

This secondary analysis study of gene expression data from a TPM clinical trial not only yielded consistent results with those of primary analysis but also identified additional new genes and pathways on TPM response to METH addiction.

Feature Subset Selection for Cancer Classification Using Weight Local Modularity

Microarray is recently becoming an important tool for profiling the global gene expression patterns of tissues. Gene selection is a popular technology for cancer classification that aims to identify a small number of informative genes from thousands of genes that may contribute to the occurrence of cancers to obtain a high predictive accuracy. This technique has been extensively studied in recent years. This study develops a novel feature selection (FS) method for gene subset selection by utilizing the Weight Local Modularity (WLM) in a complex network, called the WLMGS. In the proposed method, the discriminative power of gene subset is evaluated by using the weight local modularity of a weighted sample graph in the gene subset where the intra-class distance is small and the inter-class distance is large. A higher local modularity of the gene subset corresponds to a greater discriminative of the gene subset. With the use of forward search strategy, a more informative gene subset as a group can be selected for the classification process. Computational experiments show that the proposed algorithm can select a small subset of the predictive gene as a group while preserving classification accuracy.

Mining potential biomarkers associated with space flight in Caenorhabditis elegans experienced Shenzhou-8 mission with multiple feature selection techniques

To identify the potential biomarkers associated with space flight, a combined algorithm, which integrates the feature selection techniques, was used to deal with the microarray datasets of Caenorhabditis elegans obtained in the Shenzhou-8 mission. Compared with the ground control treatment, a total of 86 differentially expressed (DE) genes in responses to space synthetic environment or space radiation environment were identified by two filter methods. And then the top 30 ranking genes were selected by the random forest algorithm. Gene Ontology annotation and functional enrichment analyses showed that these genes were mainly associated with metabolism process. Furthermore, clustering analysis showed that 17 genes among these are positive, including 9 for space synthetic environment and 8 for space radiation environment only. These genes could be used as the biomarkers to reflect the space environment stresses. In addition, we also found that microgravity is the main stress factor to change the expression patterns of biomarkers for the short-duration spaceflight.

Feature selection methods for big data bioinformatics: A survey from the search perspective

This paper surveys main principles of feature selection and their recent applications in big data bioinformatics. Instead of the commonly used categorization into filter, wrapper, and embedded approaches to feature selection, we formulate feature selection as a combinatorial optimization or search problem and categorize feature selection methods into exhaustive search, heuristic search, and hybrid methods, where heuristic search methods may further be categorized into those with or without data-distilled feature ranking measures.

Nonparametric analysis of replicated microarray experiments

Microarrays are part of a new class of biotechnologies, which allow the monitoring of expression levels of thousands of genes simultaneously. In microarray data analysis, the comparison of gene expression profiles with respect to different conditions and the selection of biologically interesting genes are crucial tasks. Multivariate statistical methods have been applied to analyze these large data sets. To identify genes with altered expression under two experimental conditions, we propose a nonparametric statistical approach. Specifically, we propose estimating the distributions of a t-type statistic and its null statistic, using kernel methods. A comparison of these two distributions by means of a likelihood ratio test can identify genes with significantly changed expressions. A new method to provide more stable estimates of tail probabilities is proposed, as well as a method for the calculation of the cut-off point and the acceptance region. The methodology is applied to a leukaemia data set containing expression levels of 7129 genes, and is compared with normal mixture model and the traditional t-test.

f-divergence cutoff index to simultaneously identify differential expression in the integrated transcriptome and proteome

The ability to integrate ‘omics’ (i.e. transcriptomics and proteomics) is becoming increasingly important to the understanding of regulatory mechanisms. There are currently no tools available to identify differentially expressed genes (DEGs) across different ‘omics’ data types or multi-dimensional data including time courses. We present fCI (f-divergence Cut-out Index), a model capable of simultaneously identifying DEGs from continuous and discrete transcriptomic, proteomic and integrated proteogenomic data. We show that fCI can be used across multiple diverse sets of data and can unambiguously find genes that show functional modulation, developmental changes or misregulation. Applying fCI to several proteogenomics datasets, we identified a number of important genes that showed distinctive regulation patterns. The package fCI is available at R Bioconductor and http://software.steenlab.org/fCI/.

Finding differentially expressed genes in high dimensional data: Rank based test statistic via a distance measure

We present a rank-based test statistic for the identification of differentially expressed genes using a distance measure. The proposed test statistic is highly robust against extreme values and does not assume the distribution of parent population. Simulation studies show that the proposed test is more powerful than some of the commonly used methods, such as paired t-test, Wilcoxon signed rank test, and significance analysis of microarray (SAM) under certain non-normal distributions. The asymptotic distribution of the test statistic, and the p-value function are discussed. The application of proposed method is shown using a real-life data set.

A Bayesian approach for inducing sparsity in generalized linear models with multi-category response

BACKGROUND

The dimension and complexity of high-throughput gene expression data create many challenges for downstream analysis. Several approaches exist to reduce the number of variables with respect to small sample sizes. In this study, we utilized the Generalized Double Pareto (GDP) prior to induce sparsity in a Bayesian Generalized Linear Model (GLM) setting. The approach was evaluated using a publicly available microarray dataset containing 99 samples corresponding to four different prostate cancer subtypes.

RESULTS

A hierarchical Sparse Bayesian GLM using GDP prior (SBGG) was developed to take into account the progressive nature of the response variable. We obtained an average overall classification accuracy between 82.5% and 94%, which was higher than Support Vector Machine, Random Forest or a Sparse Bayesian GLM using double exponential priors. Additionally, SBGG outperforms the other 3 methods in correctly identifying pre-metastatic stages of cancer progression, which can prove extremely valuable for therapeutic and diagnostic purposes. Importantly, using Geneset Cohesion Analysis Tool, we found that the top 100 genes produced by SBGG had an average functional cohesion p-value of 2.0E-4 compared to 0.007 to 0.131 produced by the other methods.

CONCLUSIONS

Using GDP in a Bayesian GLM model applied to cancer progression data results in better subclass prediction. In particular, the method identifies pre-metastatic stages of prostate cancer with substantially better accuracy and produces more functionally relevant gene sets.

Recurrent deletions of IKZF1 in pediatric acute myeloid leukemia

IKAROS family zinc finger 1/IKZF1 is a transcription factor important in lymphoid differentiation, and a known tumor suppressor in acute lymphoid leukemia. Recent studies suggest that IKZF1 is also involved in myeloid differentiation. To investigate whether IKZF1 deletions also play a role in pediatric acute myeloid leukemia, we screened a panel of pediatric acute myeloid leukemia samples for deletions of the IKZF1 locus using multiplex ligation-dependent probe amplification and for mutations using direct sequencing. Three patients were identified with a single amino acid variant without change of IKZF1 length. No frame-shift mutations were found. Out of 11 patients with an IKZF1 deletion, 8 samples revealed a complete loss of chromosome 7, and 3 cases a focal deletion of 0.1-0.9Mb. These deletions included the complete IKZF1 gene (n=2) or exons 1-4 (n=1), all leading to a loss of IKZF1 function. Interestingly, differentially expressed genes in monosomy 7 cases (n=8) when compared to non-deleted samples (n=247) significantly correlated with gene expression changes in focal IKZF1-deleted cases (n=3). Genes with increased expression included genes involved in myeloid cell self-renewal and cell cycle, and a significant portion of GATA target genes and GATA factors. Together, these results suggest that loss of IKZF1 is recurrent in pediatric acute myeloid leukemia and might be a determinant of oncogenesis in acute myeloid leukemia with monosomy 7.

RBioCloud: A Light-Weight Framework for Bioconductor and R-based Jobs on the Cloud

Large-scale ad hoc analytics of genomic data is popular using the R-programming language supported by over 700 software packages provided by Bioconductor. More recently, analytical jobs are benefitting from on-demand computing and storage, their scalability and their low maintenance cost, all of which are offered by the cloud. While biologists and bioinformaticists can take an analytical job and execute it on their personal workstations, it remains challenging to seamlessly execute the job on the cloud infrastructure without extensive knowledge of the cloud dashboard. How analytical jobs can not only with minimum effort be executed on the cloud, but also how both the resources and data required by the job can be managed is explored in this paper. An open-source light-weight framework for executing R-scripts using Bioconductor packages, referred to as `RBioCloud', is designed and developed. RBioCloud offers a set of simple command-line tools for managing the cloud resources, the data and the execution of the job. Three biological test cases validate the feasibility of RBioCloud. The framework is available from http://www.rbiocloud.com.

Expression profiling of microarray gene signatures in acute and chronic myeloid leukaemia in human bone marrow

BACKGROUND

Classification of cancer subtypes by means of microarray signatures is becoming increasingly difficult to ignore as a potential to transform pathological diagnosis; nonetheless, measurement of Indicator genes in routine practice appears to be arduous. In a preceding published study, we utilized real-time PCR measurement of Indicator genes in acute lymphoid leukaemia (ALL) and acute myeloid leukaemia (AML) as a way of application of microarray gene signatures. More to the point, the specificity of such genes for this distinction was investigated by their measurement in cases afflicted with chronic myeloid leukaemia (CML) and with normal bone marrow (BM).

MATERIAL AND METHOD

Mononuclear cells were sorted into unselected (total), CD34+ve, and CD34-ve fractions, mRNA globally amplified by using PolyA PCR. Moreover, the level of expression of 17 Indicator genes was identified by using real-time PCR.

RESULTS

No statistically significant difference was observed in expression for any gene among CML cases. Cyclin D3 (p≤0.04) was exclusively upregulated in CML in the CD34+ fraction, notwithstanding upregulation of HkrT-1 (p≤0.02) and fumarylacetoacetate (p≤0.03) in AML. HOXA9 experienced a non-significant upregulation in AML; however, in combination with proteoglycan 1 distinguished between AML and normal samples in the CD34- fraction in unsupervised clustering. Unsupervised clustering distinguished among AML and the other diagnostic groups.

CONCLUSION

The evidence from the present study suggests that the genes discriminatory between ALL and AML are uninformative in the context of CML and normal BM, excepting for distinction with AML.

Supervised multi-view canonical correlation analysis (sMVCCA): integrating histologic and proteomic features for predicting recurrent prostate cancer

In this work, we present a new methodology to facilitate prediction of recurrent prostate cancer (CaP) following radical prostatectomy (RP) via the integration of quantitative image features and protein expression in the excised prostate. Creating a fused predictor from high-dimensional data streams is challenging because the classifier must 1) account for the "curse of dimensionality" problem, which hinders classifier performance when the number of features exceeds the number of patient studies and 2) balance potential mismatches in the number of features across different channels to avoid classifier bias towards channels with more features. Our new data integration methodology, supervised Multi-view Canonical Correlation Analysis (sMVCCA), aims to integrate infinite views of highdimensional data to provide more amenable data representations for disease classification. Additionally, we demonstrate sMVCCA using Spearman's rank correlation which, unlike Pearson's correlation, can account for nonlinear correlations and outliers. Forty CaP patients with pathological Gleason scores 6-8 were considered for this study. 21 of these men revealed biochemical recurrence (BCR) following RP, while 19 did not. For each patient, 189 quantitative histomorphometric attributes and 650 protein expression levels were extracted from the primary tumor nodule. The fused histomorphometric/proteomic representation via sMVCCA combined with a random forest classifier predicted BCR with a mean AUC of 0.74 and a maximum AUC of 0.9286. We found sMVCCA to perform statistically significantly (p < 0.05) better than comparative state-of-the-art data fusion strategies for predicting BCR. Furthermore, Kaplan-Meier analysis demonstrated improved BCR-free survival prediction for the sMVCCA-fused classifier as compared to histology or proteomic features alone.

Core and specific network markers of carcinogenesis from multiple cancer samples

Cancer is the leading cause of death worldwide and is generally caused by mutations in multiple proteins or the dysregulation of pathways. Understanding the causes and the underlying carcinogenic mechanisms can help fight this disease. In this study, a systems biology approach was used to construct the protein-protein interaction (PPI) networks of four cancers and the non-cancers by their corresponding microarray data, PPI modeling and database-mining. By comparing PPI networks between cancer and non-cancer samples to find significant proteins with large PPI changes during carcinogenesis process, core and specific network markers were identified by the intersection and difference of significant proteins, respectively, with carcinogenesis relevance values (CRVs) for each cancer. A total of 28 significant proteins were identified as core network markers in the carcinogenesis of four types of cancer, two of which are novel cancer-related proteins (e.g., UBC and PSMA3). Moreover, seven crucial common pathways were found among these cancers based on their core network markers, and some specific pathways were particularly prominent based on the specific network markers of different cancers (e.g., the RIG-I-like receptor pathway in bladder cancer, the proteasome pathway and TCR pathway in liver cancer, and the HR pathway in lung cancer). Additional validation of these network markers using the literature and new tested datasets could strengthen our findings and confirm the proposed method. From these core and specific network markers, we could not only gain an insight into crucial common and specific pathways in the carcinogenesis, but also obtain a high promising PPI target for cancer therapy.

Design and analysis of a Petri net model of the Von Hippel-Lindau (VHL) tumor suppressor interaction network

Von Hippel-Lindau (VHL) syndrome is a hereditary condition predisposing to the development of different cancer forms, related to germline inactivation of the homonymous tumor suppressor pVHL. The best characterized function of pVHL is the ubiquitination dependent degradation of Hypoxia Inducible Factor (HIF) via the proteasome. It is also involved in several cellular pathways acting as a molecular hub and interacting with more than 200 different proteins. Molecular details of pVHL plasticity remain in large part unknown. Here, we present a novel manually curated Petri Net (PN) model of the main pVHL functional pathways. The model was built using functional information derived from the literature. It includes all major pVHL functions and is able to credibly reproduce VHL syndrome at the molecular level. The reliability of the PN model also allowed in silico knockout experiments, driven by previous model analysis. Interestingly, PN analysis suggests that the variability of different VHL manifestations is correlated with the concomitant inactivation of different metabolic pathways.

Strategies for measurement of biotransformation enzyme gene expression

The analysis of gene expression is an integral part of any gene function research. A wide variety of techniques have been developed for this purpose, each with its own advantages and limitations. The following chapter seeks to provide an overview of some of the most recent as well as conventional methods to study gene expression. These approaches include Northern blot analysis, ribonuclease protection assay, reverse transcription polymerase chain reaction, expressed tag sequencing, differential display, cDNA arrays, serial analysis of gene expression, and transcriptome sequencing. The current applications of the information derived from gene expression studies require most of the assays to be adaptable for the quantitative analysis of a large number of samples and endpoints within a short period of time coupled with cost-effectiveness. A comparison of some of these features of each analytical approach as well as their advantages and disadvantages has also been provided.

DNA arrays as diagnostic tools in human healthcare

The sequencing of the entire human genome was completed in June 2000. The sequence, however, is only a starting point and gene-function is now of major interest. All this information shows that gene-based diagnostics can be helpful for treatment targeting and patient surveillance. High volume gene expression assays can optimize pharmaceutical therapies by targeting genome-based treatments to specific patient populations and providing methods to study genes involved with cancer growth patterns and tumor suppression. Molecular biology, so far, has elucidated many of the genetic mechanisms underlying heritable metabolic diseases, so that appropriate diagnostic assays will revolutionize molecular diagnostics in medicine and pharmaceuticals. One of the most promising new technologies designed to analyze large amounts of genomic information rapidly is the DNA chip.

Detecting Differentially Coexpressed Genes from Labeled Expression Data: A Brief Review

We review methods for capturing differential coexpression, which can be divided into two cases by the size of gene sets: 1) two paired genes and 2) multiple genes. In the first case, two genes are positively and negatively correlated with each other under one and the other conditions, respectively. In the second case, multiple genes are coexpressed and randomly expressed under one and the other conditions, respectively. We summarize a variety of methods for the first and second cases into four and three approaches, respectively. We describe each of these approaches in detail technically, being followed by thorough comparative experiments with both synthetic and real data sets. Our experimental results imply high possibility of improving the efficiency of the current methods, particularly in the case of multiple genes, because of low performance achieved by the best methods which are relatively simple intuitive ones.