Impact of RNA-seq data analysis algorithms on gene expression estimation and downstream prediction.

To use next-generation sequencing technology such as RNA-seq for medical and health applications, choosing proper analysis methods for biomarker identification remains a critical challenge for most users. The US Food and Drug Administration (FDA) has led the Sequencing Quality Control (SEQC) project to conduct a comprehensive investigation of 278 representative RNA-seq data analysis pipelines consisting of 13 sequence mapping, three quantification, and seven normalization methods. In this article, we focused on the impact of the joint effects of RNA-seq pipelines on gene expression estimation as well as the downstream prediction of disease outcomes. First, we developed and applied three metrics (i.e., accuracy, precision, and reliability) to quantitatively evaluate each pipeline's performance on gene expression estimation. We then investigated the correlation between the proposed metrics and the downstream prediction performance using two real-world cancer datasets (i.e., SEQC neuroblastoma dataset and the NIH/NCI TCGA lung adenocarcinoma dataset). We found that RNA-seq pipeline components jointly and significantly impacted the accuracy of gene expression estimation, and its impact was extended to the downstream prediction of these cancer outcomes. Specifically, RNA-seq pipelines that produced more accurate, precise, and reliable gene expression estimation tended to perform better in the prediction of disease outcome. In the end, we provided scenarios as guidelines for users to use these three metrics to select sensible RNA-seq pipelines for the improved accuracy, precision, and reliability of gene expression estimation, which lead to the improved downstream gene expression-based prediction of disease outcome.

Discovery of Lipidome Alterations Following Traumatic Brain Injury via High-Resolution Metabolomics.

Traumatic brain injury (TBI) can occur across wide segments of the population, presenting in a heterogeneous manner that makes diagnosis inconsistent and management challenging. Biomarkers offer the potential to objectively identify injury status, severity, and phenotype by measuring the relative concentrations of endogenous molecules in readily accessible biofluids. Through a data-driven, discovery approach, novel biomarker candidates for TBI were identified in the serum lipidome of adult male Sprague-Dawley rats in the first week following moderate controlled cortical impact (CCI). Serum samples were analyzed in positive and negative modes by ultraperformance liquid chromatography-mass spectrometry (UPLC-MS). A predictive panel for the classification of injured and uninjured sera samples, consisting of 26 dysregulated species belonging to a variety of lipid classes, was developed with a cross-validated accuracy of 85.3% using omniClassifier software to optimize feature selection. Polyunsaturated fatty acids (PUFAs) and PUFA-containing diacylglycerols were found to be upregulated in sera from injured rats, while changes in sphingolipids and other membrane phospholipids were also observed, many of which map to known secondary injury pathways. Overall, the identified biomarker panel offers viable molecular candidates representing lipids that may readily cross the blood-brain barrier (BBB) and aid in the understanding of TBI pathophysiology.

Integration of Multi-Modal Biomedical Data to Predict Cancer Grade and Patient Survival.

The Big Data era in Biomedical research has resulted in large-cohort data repositories such as The Cancer Genome Atlas (TCGA). These repositories routinely contain hundreds of matched patient samples for genomic, proteomic, imaging, and clinical data modalities, enabling holistic and multi-modal integrative analysis of human disease. Using TCGA renal and ovarian cancer data, we conducted a novel investigation of multi-modal data integration by combining histopathological image and RNA-seq data. We compared the performances of two integrative prediction methods: majority vote and stacked generalization. Results indicate that integration of multiple data modalities improves prediction of cancer grade and outcome. Specifically, stacked generalization, a method that integrates multiple data modalities to produce a single prediction result, outperforms both single-data-modality prediction and majority vote. Moreover, stacked generalization reveals the contribution of each data modality (and specific features within each data modality) to the final prediction result and may provide biological insights to explain prediction performance.

A Multi-Modal Graph-Based Semi-Supervised Pipeline for Predicting Cancer Survival.

Cancer survival prediction is an active area of research that can help prevent unnecessary therapies and improve patient's quality of life. Gene expression profiling is being widely used in cancer studies to discover informative biomarkers that aid predict different clinical endpoint prediction. We use multiple modalities of data derived from RNA deep-sequencing (RNA-seq) to predict survival of cancer patients. Despite the wealth of information available in expression profiles of cancer tumors, fulfilling the aforementioned objective remains a big challenge, for the most part, due to the paucity of data samples compared to the high dimension of the expression profiles. As such, analysis of transcriptomic data modalities calls for state-of-the-art big-data analytics techniques that can maximally use all the available data to discover the relevant information hidden within a significant amount of noise. In this paper, we propose a pipeline that predicts cancer patients' survival by exploiting the structure of the input (manifold learning) and by leveraging the unlabeled samples using Laplacian support vector machines, a graph-based semi supervised learning (GSSL) paradigm. We show that under certain circumstances, no single modality per se will result in the best accuracy and by fusing different models together via a stacked generalization strategy, we may boost the accuracy synergistically. We apply our approach to two cancer datasets and present promising results. We maintain that a similar pipeline can be used for predictive tasks where labeled samples are expensive to acquire.

Comparison of RNA-seq and microarray-based models for clinical endpoint prediction.

BACKGROUND: Gene expression profiling is being widely applied in cancer research to identify biomarkers for clinical endpoint prediction. Since RNA-seq provides a powerful tool for transcriptome-based applications beyond the limitations of microarrays, we sought to systematically evaluate the performance of RNA-seq-based and microarray-based classifiers in this MAQC-III/SEQC study for clinical endpoint prediction using neuroblastoma as a model. RESULTS: We generate gene expression profiles from 498 primary neuroblastomas using both RNA-seq and 44 k microarrays. Characterization of the neuroblastoma transcriptome by RNA-seq reveals that more than 48,000 genes and 200,000 transcripts are being expressed in this malignancy. We also find that RNA-seq provides much more detailed information on specific transcript expression patterns in clinico-genetic neuroblastoma subgroups than microarrays. To systematically compare the power of RNA-seq and microarray-based models in predicting clinical endpoints, we divide the cohort randomly into training and validation sets and develop 360 predictive models on six clinical endpoints of varying predictability. Evaluation of factors potentially affecting model performances reveals that prediction accuracies are most strongly influenced by the nature of the clinical endpoint, whereas technological platforms (RNA-seq vs. microarrays), RNA-seq data analysis pipelines, and feature levels (gene vs. transcript vs. exon-junction level) do not significantly affect performances of the models. CONCLUSIONS: We demonstrate that RNA-seq outperforms microarrays in determining the transcriptomic characteristics of cancer, while RNA-seq and microarray-based models perform similarly in clinical endpoint prediction. Our findings may be valuable to guide future studies on the development of gene expression-based predictive models and their implementation in clinical practice.

Effect of low-expression gene filtering on detection of differentially expressed genes in RNA-seq data.

We compare methods for filtering RNA-seq lowexpression genes and investigate the effect of filtering on detection of differentially expressed genes (DEGs). Although RNA-seq technology has improved the dynamic range of gene expression quantification, low-expression genes may be indistinguishable from sampling noise. The presence of noisy, low-expression genes can decrease the sensitivity of detecting DEGs. Thus, identification and filtering of these low-expression genes may improve DEG detection sensitivity. Using the SEQC benchmark dataset, we investigate the effect of different filtering methods on DEG detection sensitivity. Moreover, we investigate the effect of RNA-seq pipelines on optimal filtering thresholds. Results indicate that the filtering threshold that maximizes the total number of DEGs closely corresponds to the threshold that maximizes DEG detection sensitivity. Transcriptome reference annotation, expression quantification method, and DEG detection method are statistically significant RNA-seq pipeline factors that affect the optimal filtering threshold.

The impact of RNA-seq aligners on gene expression estimation.

While numerous RNA-seq data analysis pipelines are available, research has shown that the choice of pipeline influences the results of differentially expressed gene detection and gene expression estimation. Gene expression estimation is a key step in RNA-seq data analysis, since the accuracy of gene expression estimates profoundly affects the subsequent analysis. Generally, gene expression estimation involves sequence alignment and quantification, and accurate gene expression estimation requires accurate alignment. However, the impact of aligners on gene expression estimation remains unclear. We address this need by constructing nine pipelines consisting of nine spliced aligners and one quantifier. We then use simulated data to investigate the impact of aligners on gene expression estimation. To evaluate alignment, we introduce three alignment performance metrics, (1) the percentage of reads aligned, (2) the percentage of reads aligned with zero mismatch (ZeroMismatchPercentage), and (3) the percentage of reads aligned with at most one mismatch (ZeroOneMismatchPercentage). We then evaluate the impact of alignment performance on gene expression estimation using three metrics, (1) gene detection accuracy, (2) the number of genes falsely quantified (FalseExpNum), and (3) the number of genes with falsely estimated fold changes (FalseFcNum). We found that among various pipelines, FalseExpNum and FalseFcNum are correlated. Moreover, FalseExpNum is linearly correlated with the percentage of reads aligned and ZeroMismatchPercentage, and FalseFcNum is linearly correlated with ZeroMismatchPercentage. Because of this correlation, the percentage of reads aligned and ZeroMismatchPercentage may be used to assess the performance of gene expression estimation for all RNA-seq datasets.

A semi-supervised method for predicting cancer survival using incomplete clinical data.

Prediction of survival for cancer patients is an open area of research. However, many of these studies focus on datasets with a large number of patients. We present a novel method that is specifically designed to address the challenge of data scarcity, which is often the case for cancer datasets. Our method is able to use unlabeled data to improve classification by adopting a semi-supervised training approach to learn an ensemble classifier. The results of applying our method to three cancer datasets show the promise of semi-supervised learning for prediction of cancer survival.

Detection of blur artifacts in histopathological whole-slide images of endomyocardial biopsies.

Histopathological whole-slide images (WSIs) have emerged as an objective and quantitative means for image-based disease diagnosis. However, WSIs may contain acquisition artifacts that affect downstream image feature extraction and quantitative disease diagnosis. We develop a method for detecting blur artifacts in WSIs using distributions of local blur metrics. As features, these distributions enable accurate classification of WSI regions as sharp or blurry. We evaluate our method using over 1000 portions of an endomyocardial biopsy (EMB) WSI. Results indicate that local blur metrics accurately detect blurry image regions.

Removing batch effects from histopathological images for enhanced cancer diagnosis.

Researchers have developed computer-aided decision support systems for translational medicine that aim to objectively and efficiently diagnose cancer using histopathological images. However, the performance of such systems is confounded by nonbiological experimental variations or "batch effects" that can commonly occur in histopathological data, especially when images are acquired using different imaging devices and patient samples. This is even more problematic in large-scale studies in which cross-laboratory sharing of large volumes of data is necessary. Batch effects can change quantitative morphological image features and decrease the prediction performance. Using four batches of renal tumor images, we compare one image-level and five feature-level batch effect removal methods. Principal component variation analysis shows that batch is a large source of variance in image features. Results show that feature-level normalization methods reduce batch-contributed variance to almost zero. Moreover, feature-level normalization, especially ComBatN, improves cross-batch and combined-batch prediction performance. Compared to no normalization, ComBatN improves performance in 83% and 90% of cross-batch and combined-batch prediction models, respectively.

Investigation of factors affecting RNA-seq gene expression calls.

RNA-seq enables quantification of the human transcriptome. Estimation of gene expression is a fundamental issue in the analysis of RNA-seq data. However, there is an inherent ambiguity in distinguishing between genes with very low expression and experimental or transcriptional noise. We conducted an exploratory investigation of some factors that may affect gene expression calls. We observed that the distribution of reads that map to exonic, intronic, and intergenic regions are distinct. These distributions may provide useful insights into the behavior of gene expression noise. Moreover, we observed that these distributions are qualitatively similar between two sequence mapping algorithms. Finally, we examined the relationship between gene length and gene expression calls, and observed that they are correlated. This preliminary investigation is important for RNA-seq gene expression analysis because it may lead to more effective algorithms for distinguishing between true gene expression and experimental or transcriptional noise.

omniClassifier: a Desktop Grid Computing System for Big Data Prediction Modeling.

Robust prediction models are important for numerous science, engineering, and biomedical applications. However, best-practice procedures for optimizing prediction models can be computationally complex, especially when choosing models from among hundreds or thousands of parameter choices. Computational complexity has further increased with the growth of data in these fields, concurrent with the era of "Big Data". Grid computing is a potential solution to the computational challenges of Big Data. Desktop grid computing, which uses idle CPU cycles of commodity desktop machines, coupled with commercial cloud computing resources can enable research labs to gain easier and more cost effective access to vast computing resources. We have developed omniClassifier, a multi-purpose prediction modeling application that provides researchers with a tool for conducting machine learning research within the guidelines of recommended best-practices. omniClassifier is implemented as a desktop grid computing system using the Berkeley Open Infrastructure for Network Computing (BOINC) middleware. In addition to describing implementation details, we use various gene expression datasets to demonstrate the potential scalability of omniClassifier for efficient and robust Big Data prediction modeling. A prototype of omniClassifier can be accessed at http://omniclassifier.bme.gatech.edu/.

The Impact of RNA-seq Alignment Pipeline on Detection of Differentially Expressed Genes.

RNA-seq data analysis pipelines are generally composed of sequence alignment, expression quantification, expression normalization, and differentially expressed gene (DEG) detection. Each step has numerous specific tools or algorithms, so we cannot explore all combinatorial pipelines and provide a comprehensive comparison of pipeline performance. To understand the mechanism of RNA-seq data analysis pipelines and provide some useful information for pipeline selection, we believe it is necessary to analyze the interactions among pipeline components. In this paper, by combining different alignment algorithms with the same quantification, normalization, and DEG detection tools, we construct nine RNA-seq pipelines to analyze the impact of RNA-seq alignment on downstream applications of gene expression estimates. Specifically, we find moderate linear correlation between the number of DEGs detected and the percentage of reads aligned with zero mismatch.

Benchmarking RNA-Seq quantification tools.

RNA-Seq, a deep sequencing technique, promises to be a potential successor to microarrays for studying the transcriptome. One of many aspects of transcriptomics that are of interest to researchers is gene expression estimation. With rapid development in RNA-Seq, there are numerous tools available to estimate gene expression, each producing different results. However, we do not know which of these tools produces the most accurate gene expression estimates. In this study we have addressed this issue using Cufflinks, IsoEM, HTSeq, and RSEM to quantify RNA-Seq expression profiles. Comparing results of these quantification tools, we observe that RNA-Seq relative expression estimates correlate with RT-qPCR measurements in the range of 0.85 to 0.89, with HTSeq exhibiting the highest correlation. But, in terms of root-mean-square deviation of RNA-Seq relative expression estimates from RT-qPCR measurements, we find HTSeq to produce the greatest deviation. Therefore, we conclude that, though Cufflinks, RSEM, and IsoEM might not correlate as well as HTSeq with RT-qPCR measurements, they may produce expression values with higher accuracy.

Eliminating tissue-fold artifacts in histopathological whole-slide images for improved image-based prediction of cancer grade.

BACKGROUND: Analysis of tissue biopsy whole-slide images (WSIs) depends on effective detection and elimination of image artifacts. We present a novel method to detect tissue-fold artifacts in histopathological WSIs. We also study the effect of tissue folds on image features and prediction models. MATERIALS AND METHODS: We use WSIs of samples from two cancer endpoints - kidney clear cell carcinoma (KiCa) and ovarian serous adenocarcinoma (OvCa) - publicly available from The Cancer Genome Atlas. We detect tissue folds in low-resolution WSIs using color properties and two adaptive connectivity-based thresholds. We optimize and validate our tissue-fold detection method using 105 manually annotated WSIs from both cancer endpoints. In addition to detecting tissue folds, we extract 461 image features from the high-resolution WSIs for all samples. We use the rank-sum test to find image features that are statistically different among features extracted from the same set of WSIs with and without folds. We then use features that are affected by tissue folds to develop models for predicting cancer grades. RESULTS: When compared to the ground truth, our method detects tissue folds in KiCa with 0.50 adjusted Rand index (ARI), 0.77 average true rate (ATR), 0.55 true positive rate (TPR), and 0.98 true negative rate (TNR); and in OvCa with 0.40 ARI, 0.73 ATR, 0.47 TPR, and 0.98 TNR. Compared to two other methods, our method is more accurate in terms of ARI and ATR. We found that 53 and 30 image features were significantly affected by the presence of tissue-fold artifacts (detected using our method) in OvCa and KiCa, respectively. After eliminating tissue folds, the performance of cancer-grade prediction models improved by 5% and 1% in OvCa and KiCa, respectively. CONCLUSION: The proposed connectivity-based method is more effective in detecting tissue folds compared to other methods. Reducing tissue-fold artifacts will increase the performance of cancer-grade prediction models.

Computational nanomedicine: modeling of nanoparticle-mediated hyperthermal cancer therapy.

Nanoparticle-mediated hyperthermia for cancer therapy is a growing area of cancer nanomedicine because of the potential for localized and targeted destruction of cancer cells. Localized hyperthermal effects are dependent on many factors, including nanoparticle size and shape, excitation wavelength and power, and tissue properties. Computational modeling is an important tool for investigating and optimizing these parameters. In this review, we focus on computational modeling of magnetic and gold nanoparticle-mediated hyperthermia, followed by a discussion of new opportunities and challenges.

Pathology imaging informatics for quantitative analysis of whole-slide images.

OBJECTIVES: With the objective of bringing clinical decision support systems to reality, this article reviews histopathological whole-slide imaging informatics methods, associated challenges, and future research opportunities. TARGET AUDIENCE: This review targets pathologists and informaticians who have a limited understanding of the key aspects of whole-slide image (WSI) analysis and/or a limited knowledge of state-of-the-art technologies and analysis methods. SCOPE: First, we discuss the importance of imaging informatics in pathology and highlight the challenges posed by histopathological WSI. Next, we provide a thorough review of current methods for: quality control of histopathological images; feature extraction that captures image properties at the pixel, object, and semantic levels; predictive modeling that utilizes image features for diagnostic or prognostic applications; and data and information visualization that explores WSI for de novo discovery. In addition, we highlight future research directions and discuss the impact of large public repositories of histopathological data, such as the Cancer Genome Atlas, on the field of pathology informatics. Following the review, we present a case study to illustrate a clinical decision support system that begins with quality control and ends with predictive modeling for several cancer endpoints. Currently, state-of-the-art software tools only provide limited image processing capabilities instead of complete data analysis for clinical decision-making. We aim to inspire researchers to conduct more research in pathology imaging informatics so that clinical decision support can become a reality.

An integrative approach for the large-scale identification of human genome kinases regulating cancer metastasis.

Kinases become one of important groups of drug targets. To identify more kinases being potential for cancer therapy, we developed an integrative approach for the large-scale screen of functional genes capable of regulating the main traits of cancer metastasis. We first employed self-assembled cell microarray to screen functional genes that regulate cancer cell migration using a human genome kinase siRNA library. We identified 81 genes capable of significantly regulating cancer cell migration. Following with invasion assays and bio-informatics analysis, we discovered that 16 genes with differentially expression in cancer samples can regulate both cell migration and invasion, among which 10 genes have been well known to play critical roles in the cancer development. The remaining 6 genes were experimentally validated to have the capacities of regulating cell proliferation, apoptosis and anoikis activities besides cell motility. Together, these findings provide a new insight into the therapeutic use of human kinases. FROM THE CLINICAL EDITOR: This team of authors have utilized a self-assembled cell microarray to screen genes that regulate cancer cell migration using a human genome siRNA library of kinases. They validated previously known genes and identified novel ones that may serve as therapeutic targets.

Histological image classification using biologically interpretable shape-based features.

BACKGROUND: Automatic cancer diagnostic systems based on histological image classification are important for improving therapeutic decisions. Previous studies propose textural and morphological features for such systems. These features capture patterns in histological images that are useful for both cancer grading and subtyping. However, because many of these features lack a clear biological interpretation, pathologists may be reluctant to adopt these features for clinical diagnosis. METHODS: We examine the utility of biologically interpretable shape-based features for classification of histological renal tumor images. Using Fourier shape descriptors, we extract shape-based features that capture the distribution of stain-enhanced cellular and tissue structures in each image and evaluate these features using a multi-class prediction model. We compare the predictive performance of the shape-based diagnostic model to that of traditional models, i.e., using textural, morphological and topological features. RESULTS: The shape-based model, with an average accuracy of 77%, outperforms or complements traditional models. We identify the most informative shapes for each renal tumor subtype from the top-selected features. Results suggest that these shapes are not only accurate diagnostic features, but also correlate with known biological characteristics of renal tumors. CONCLUSIONS: Shape-based analysis of histological renal tumor images accurately classifies disease subtypes and reveals biologically insightful discriminatory features. This method for shape-based analysis can be extended to other histological datasets to aid pathologists in diagnostic and therapeutic decisions.