Google goes cancer: improving outcome prediction for cancer patients by network-based ranking of marker genes

Predicting the clinical outcome of cancer patients based on the expression of marker genes in their tumors has received increasing interest in the past decade. Accurate predictors of outcome and response to therapy could be used to personalize and thereby improve therapy. However, state of the art methods used so far often found marker genes with limited prediction accuracy, limited reproducibility, and unclear biological relevance. To address this problem, we developed a novel computational approach to identify genes prognostic for outcome that couples gene expression measurements from primary tumor samples with a network of known relationships between the genes. Our approach ranks genes according to their prognostic relevance using both expression and network information in a manner similar to Google's PageRank. We applied this method to gene expression profiles which we obtained from 30 patients with pancreatic cancer, and identified seven candidate marker genes prognostic for outcome. Compared to genes found with state of the art methods, such as Pearson correlation of gene expression with survival time, we improve the prediction accuracy by up to 7%. Accuracies were assessed using support vector machine classifiers and Monte Carlo cross-validation. We then validated the prognostic value of our seven candidate markers using immunohistochemistry on an independent set of 412 pancreatic cancer samples. Notably, signatures derived from our candidate markers were independently predictive of outcome and superior to established clinical prognostic factors such as grade, tumor size, and nodal status. As the amount of genomic data of individual tumors grows rapidly, our algorithm meets the need for powerful computational approaches that are key to exploit these data for personalized cancer therapies in clinical practice.     

Ensemble Positive Unlabeled Learning for Disease Gene Identification

An increasing number of genes have been experimentally confirmed in recent years as causative genes to various human diseases. The newly available knowledge can be exploited by machine learning methods to discover additional unknown genes that are likely to be associated with diseases. In particular, positive unlabeled learning (PU learning) methods, which require only a positive training set P (confirmed disease genes) and an unlabeled set U (the unknown candidate genes) instead of a negative training set N, have been shown to be effective in uncovering new disease genes in the current scenario. Using only a single source of data for prediction can be susceptible to bias due to incompleteness and noise in the genomic data and a single machine learning predictor prone to bias caused by inherent limitations of individual methods. In this paper, we propose an effective PU learning framework that integrates multiple biological data sources and an ensemble of powerful machine learning classifiers for disease gene identification. Our proposed method integrates data from multiple biological sources for training PU learning classifiers. A novel ensemble-based PU learning method EPU is then used to integrate multiple PU learning classifiers to achieve accurate and robust disease gene predictions. Our evaluation experiments across six disease groups showed that EPU achieved significantly better results compared with various state-of-the-art prediction methods as well as ensemble learning classifiers. Through integrating multiple biological data sources for training and the outputs of an ensemble of PU learning classifiers for prediction, we are able to minimize the potential bias and errors in individual data sources and machine learning algorithms to achieve more accurate and robust disease gene predictions. In the future, our EPU method provides an effective framework to integrate the additional biological and computational resources for better disease gene predictions.     

Global risk transformative prioritization for prostate cancer candidate genes in molecular networks

Understanding the pathogenesis of complex diseases is aided by precise identification of the genes responsible. Many computational methods have been developed to prioritize candidate disease genes, but coverage of functional annotations may be a limiting factor for most of these methods. Here, we introduce a global candidate gene prioritization approach that considers information about network properties in the human protein interaction network and risk transformative contents from known disease genes. Global risk transformative scores were then used to prioritize candidate genes. This method was introduced to prioritize candidate genes for prostate cancer. The effectiveness of our global risk transformative algorithm for prioritizing candidate genes was evaluated according to validation studies. Compared with ToppGene and random walk-based methods, our method outperformed the two other candidate gene prioritization methods. The generality of our method was assessed by testing it on prostate cancer and other types of cancer. The performance was evaluated using standard leave-one-out cross-validation.     

A critical evaluation of network and pathway-based classifiers for outcome prediction in breast cancer

Recently, several classifiers that combine primary tumor data, like gene expression data, and secondary data sources, such as protein-protein interaction networks, have been proposed for predicting outcome in breast cancer. In these approaches, new composite features are typically constructed by aggregating the expression levels of several genes. The secondary data sources are employed to guide this aggregation. Although many studies claim that these approaches improve classification performance over single genes classifiers, the gain in performance is difficult to assess. This stems mainly from the fact that different breast cancer data sets and validation procedures are employed to assess the performance. Here we address these issues by employing a large cohort of six breast cancer data sets as benchmark set and by performing an unbiased evaluation of the classification accuracies of the different approaches. Contrary to previous claims, we find that composite feature classifiers do not outperform simple single genes classifiers. We investigate the effect of (1) the number of selected features; (2) the specific gene set from which features are selected; (3) the size of the training set and (4) the heterogeneity of the data set on the performance of composite feature and single genes classifiers. Strikingly, we find that randomization of secondary data sources, which destroys all biological information in these sources, does not result in a deterioration in performance of composite feature classifiers. Finally, we show that when a proper correction for gene set size is performed, the stability of single genes sets is similar to the stability of composite feature sets. Based on these results there is currently no reason to prefer prognostic classifiers based on composite features over single genes classifiers for predicting outcome in breast cancer.     

Deep short-read sequencing of chromosome 17 from the mouse strains A/J and CAST/Ei identifies significant germline variation and candidate genes that regulate liver triglyceride levels

Genome sequences are essential tools for comparative and mutational analyses. Here we present the short read sequence of mouse chromosome 17 from the Mus musculus domesticus derived strain A/J, and the Mus musculus castaneus derived strain CAST/Ei. We describe approaches for the accurate identification of nucleotide and structural variation in the genomes of vertebrate experimental organisms, and show how these techniques can be applied to help prioritize candidate genes within quantitative trait loci.     

Analyzing Large Gene Expression and Methylation Data Profiles Using StatBicRM: Statistical Biclustering-Based Rule Mining

Microarray and beadchip are two most efficient techniques for measuring gene expression and methylation data in bioinformatics. Biclustering deals with the simultaneous clustering of genes and samples. In this article, we propose a computational rule mining framework, StatBicRM (i.e., statistical biclustering-based rule mining) to identify special type of rules and potential biomarkers using integrated approaches of statistical and binary inclusion-maximal biclustering techniques from the biological datasets. At first, a novel statistical strategy has been utilized to eliminate the insignificant/low-significant/redundant genes in such way that significance level must satisfy the data distribution property (viz., either normal distribution or non-normal distribution). The data is then discretized and post-discretized, consecutively. Thereafter, the biclustering technique is applied to identify maximal frequent closed homogeneous itemsets. Corresponding special type of rules are then extracted from the selected itemsets. Our proposed rule mining method performs better than the other rule mining algorithms as it generates maximal frequent closed homogeneous itemsets instead of frequent itemsets. Thus, it saves elapsed time, and can work on big dataset. Pathway and Gene Ontology analyses are conducted on the genes of the evolved rules using David database. Frequency analysis of the genes appearing in the evolved rules is performed to determine potential biomarkers. Furthermore, we also classify the data to know how much the evolved rules are able to describe accurately the remaining test (unknown) data. Subsequently, we also compare the average classification accuracy, and other related factors with other rule-based classifiers. Statistical significance tests are also performed for verifying the statistical relevance of the comparative results. Here, each of the other rule mining methods or rule-based classifiers is also starting with the same post-discretized data-matrix. Finally, we have also included the integrated analysis of gene expression and methylation for determining epigenetic effect (viz., effect of methylation) on gene expression level.     

Design and analysis issues in genome-wide somatic mutation studies of cancer

The availability of the human genome sequence and progress in sequencing and bioinformatic technologies have enabled genome-wide investigation of somatic mutations in human cancers. This article briefly reviews challenges arising in the statistical analysis of mutational data of this kind. A first challenge is that of designing studies that efficiently allocate sequencing resources. We show that this can be addressed by two-stage designs and demonstrate via simulations that even relatively small studies can produce lists of candidate cancer genes that are highly informative for future research efforts. A second challenge is to distinguish mutated genes that are selected for by cancer (drivers) from mutated genes that have no role in the development of cancer and simply happened to mutate (passengers). We suggest that this question is best approached as a classification problem and discuss some of the difficulties of more traditional testing-based approaches. A third challenge is to identify biologic processes affected by the driver genes. This can be pursued by gene set analyses. These can reliably identify functional groups and pathways that are enriched for mutated genes even when the individual genes involved in those pathways or sets are not mutated at sufficient frequencies to provide conclusive evidence as drivers.     

An Assessment of Heavy Ion Irradiation Mutagenesis for Reverse Genetics in Wheat (Triticum aestivum L.)

Reverse genetic techniques harnessing mutational approaches are powerful tools that can provide substantial insight into gene function in plants. However, as compared to diploid species, reverse genetic analyses in polyploid plants such as bread wheat can present substantial challenges associated with high levels of sequence and functional similarity amongst homoeologous loci. We previously developed a high-throughput method to identify deletions of genes within a physically mutagenized wheat population. Here we describe our efforts to combine multiple homoeologous deletions of three candidate disease susceptibility genes (TaWRKY11, TaPFT1 and TaPLDß1). We were able to produce lines featuring homozygous deletions at two of the three homoeoloci for all genes, but this was dependent on the individual mutants used in crossing. Intriguingly, despite extensive efforts, viable lines possessing homozygous deletions at all three homoeoloci could not be produced for any of the candidate genes. To investigate deletion size as a possible reason for this phenomenon, we developed an amplicon sequencing approach based on synteny to Brachypodium distachyon to assess the size of the deletions removing one candidate gene (TaPFT1) in our mutants. These analyses revealed that genomic deletions removing the locus are relatively large, resulting in the loss of multiple additional genes. The implications of this work for the use of heavy ion mutagenesis for reverse genetic analyses in wheat are discussed.     

PanClassif: Improving pan cancer classification of single cell RNA-seq gene expression data using machine learning

Cancer is one of the major causes of human death per year. In recent years, cancer identification and classification using machine learning have gained momentum due to the availability of high throughput sequencing data. Using RNA-seq, cancer research is blooming day by day and new insights of cancer and related treatments are coming into light. In this paper, we propose PanClassif, a method that requires a very few and effective genes to detect cancer from RNA-seq data and is able to provide performance gain in several wide range machine learning classifiers. We have taken 22 types of cancer samples from The Cancer Genome Atlas (TCGA) having 8287 cancer samples and 680 normal samples. Firstly, PanClassif uses k-Nearest Neighbour (k-NN) smoothing to smooth the samples to handle noise in the data. Then effective genes are selected by Anova based test. For balancing the train data, PanClassif applies an oversampling method, SMOTE. We have performed comprehensive experiments on the datasets using several classification algorithms. Experimental results shows that PanClassif outperform existing state-of-the-art methods available and shows consistent performance for two single cell RNA-seq datasets taken from Gene Expression Omnibus (GEO). PanClassif improves performances of a wide variety of classifiers for both binary cancer prediction and multi-class cancer classification. PanClassif is available as a python package (https://pypi.org/project/panclassif/). All the source code and materials of PanClassif are available at https://github.com/Zwei-inc/panclassif.     

Mining Proteins with Non-Experimental Annotations Based on an Active Sample Selection Strategy for Predicting Protein Subcellular Localization

Subcellular localization of a protein is important to understand proteins’ functions and interactions. There are many techniques based on computational methods to predict protein subcellular locations, but it has been shown that many prediction tasks have a training data shortage problem. This paper introduces a new method to mine proteins with non-experimental annotations, which are labeled by non-experimental evidences of protein databases to overcome the training data shortage problem. A novel active sample selection strategy is designed, taking advantage of active learning technology, to actively find useful samples from the entire data pool of candidate proteins with non-experimental annotations. This approach can adequately estimate the “value” of each sample, automatically select the most valuable samples and add them into the original training set, to help to retrain the classifiers. Numerical experiments with for four popular multi-label classifiers on three benchmark datasets show that the proposed method can effectively select the valuable samples to supplement the original training set and significantly improve the performances of predicting classifiers.     

The MORPH Algorithm: Ranking Candidate Genes for Membership in Arabidopsis and Tomato Pathways

Closing gaps in our current knowledge about biological pathways is a fundamental challenge. The development of novel computational methods along with high-throughput experimental data carries the promise to help in the challenge. We present an algorithm called MORPH (for module-guided ranking of candidate pathway genes) for revealing unknown genes in biological pathways. The method receives as input a set of known genes from the target pathway, a collection of expression profiles, and interaction and metabolic networks. Using machine learning techniques, MORPH selects the best combination of data and analysis method and outputs a ranking of candidate genes predicted to belong to the target pathway. We tested MORPH on 230 known pathways in Arabidopsis thaliana and 93 known pathways in tomato (Solanum lycopersicum) and obtained high-quality cross-validation results. In the photosynthesis light reactions, homogalacturonan biosynthesis, and chlorophyll biosynthetic pathways of Arabidopsis, genes ranked highly by MORPH were recently verified to be associated with these pathways. MORPH candidates ranked for the carotenoid pathway from Arabidopsis and tomato are derived from pathways that compete for common precursors or from pathways that are coregulated with or regulate the carotenoid biosynthetic pathway.     

Simple decision rules for classifying human cancers from gene expression profiles

MOTIVATION: Various studies have shown that cancer tissue samples can be successfully detected and classified by their gene expression patterns using machine learning approaches. One of the challenges in applying these techniques for classifying gene expression data is to extract accurate, readily interpretable rules providing biological insight as to how classification is performed. Current methods generate classifiers that are accurate but difficult to interpret. This is the trade-off between credibility and comprehensibility of the classifiers. Here, we introduce a new classifier in order to address these problems. It is referred to as k-TSP (k-Top Scoring Pairs) and is based on the concept of 'relative expression reversals'. This method generates simple and accurate decision rules that only involve a small number of gene-to-gene expression comparisons, thereby facilitating follow-up studies. RESULTS: In this study, we have compared our approach to other machine learning techniques for class prediction in 19 binary and multi-class gene expression datasets involving human cancers. The k-TSP classifier performs as efficiently as Prediction Analysis of Microarray and support vector machine, and outperforms other learning methods (decision trees, k-nearest neighbour and na?ve Bayes). Our approach is easy to interpret as the classifier involves only a small number of informative genes. For these reasons, we consider the k-TSP method to be a useful tool for cancer classification from microarray gene expression data. AVAILABILITY: The software and datasets are available at http://www.ccbm.jhu.edu CONTACT: actan@jhu.edu.     

Unsupervised Inference of Protein Fitness Landscape from Deep Mutational Scan

The recent technological advances underlying the screening of large combinatorial libraries in high-throughput mutational scans deepen our understanding of adaptive protein evolution and boost its applications in protein design. Nevertheless, the large number of possible genotypes requires suitable computational methods for data analysis, the prediction of mutational effects, and the generation of optimized sequences. We describe a computational method that, trained on sequencing samples from multiple rounds of a screening experiment, provides a model of the genotype–fitness relationship. We tested the method on five large-scale mutational scans, yielding accurate predictions of the mutational effects on fitness. The inferred fitness landscape is robust to experimental and sampling noise and exhibits high generalization power in terms of broader sequence space exploration and higher fitness variant predictions. We investigate the role of epistasis and show that the inferred model provides structural information about the 3D contacts in the molecular fold.     

An unbiased adaptive sampling algorithm for the exploration of RNA mutational landscapes under evolutionary pressure

The analysis of the relationship between sequences and structures (i.e., how mutations affect structures and reciprocally how structures influence mutations) is essential to decipher the principles driving molecular evolution, to infer the origins of genetic diseases, and to develop bioengineering applications such as the design of artificial molecules. Because their structures can be predicted from the sequence data only, RNA molecules provide a good framework to study this sequence-structure relationship. We recently introduced a suite of algorithms called RNAmutants which allows a complete exploration of RNA sequence-structure maps in polynomial time and space. Formally, RNAmutants takes an input sequence (or seed) to compute the Boltzmann-weighted ensembles of mutants with exactly k mutations, and sample mutations from these ensembles. However, this approach suffers from major limitations. Indeed, since the Boltzmann probabilities of the mutations depend of the free energy of the structures, RNAmutants has difficulties to sample mutant sequences with low G+C-contents. In this article, we introduce an unbiased adaptive sampling algorithm that enables RNAmutants to sample regions of the mutational landscape poorly covered by classical algorithms. We applied these methods to sample mutations with low G+C-contents. These adaptive sampling techniques can be easily adapted to explore other regions of the sequence and structural landscapes which are difficult to sample. Importantly, these algorithms come at a minimal computational cost. We demonstrate the insights offered by these techniques on studies of complete RNA sequence structures maps of sizes up to 40 nucleotides. Our results indicate that the G+C-content has a strong influence on the size and shape of the evolutionary accessible sequence and structural spaces. In particular, we show that low G+C-contents favor the apparition of internal loops and thus possibly the synthesis of tertiary structure motifs. On the other hand, high G+C-contents significantly reduce the size of the evolutionary accessible mutational landscapes.     

The CanOE strategy: integrating genomic and metabolic contexts across multiple prokaryote genomes to find candidate genes for orphan enzymes

Of all biochemically characterized metabolic reactions formalized by the IUBMB, over one out of four have yet to be associated with a nucleic or protein sequence, i.e. are sequence-orphan enzymatic activities. Few bioinformatics annotation tools are able to propose candidate genes for such activities by exploiting context-dependent rather than sequence-dependent data, and none are readily accessible and propose result integration across multiple genomes. Here, we present CanOE (Candidate genes for Orphan Enzymes), a four-step bioinformatics strategy that proposes ranked candidate genes for sequence-orphan enzymatic activities (or orphan enzymes for short). The first step locates "genomic metabolons", i.e. groups of co-localized genes coding proteins catalyzing reactions linked by shared metabolites, in one genome at a time. These metabolons can be particularly helpful for aiding bioanalysts to visualize relevant metabolic data. In the second step, they are used to generate candidate associations between un-annotated genes and gene-less reactions. The third step integrates these gene-reaction associations over several genomes using gene families, and summarizes the strength of family-reaction associations by several scores. In the final step, these scores are used to rank members of gene families which are proposed for metabolic reactions. These associations are of particular interest when the metabolic reaction is a sequence-orphan enzymatic activity. Our strategy found over 60,000 genomic metabolons in more than 1,000 prokaryote organisms from the MicroScope platform, generating candidate genes for many metabolic reactions, of which more than 70 distinct orphan reactions. A computational validation of the approach is discussed. Finally, we present a case study on the anaerobic allantoin degradation pathway in Escherichia coli K-12.