Search

The human microbiome is comprised of thousands of microbial species. These species will substantially influence the normal physiology of humans and cause numerous diseases. Microbiome data can be measured by sequencing, microarray, or other technologies. With the fast development of these technologies, downstream analysis methods should also be designed to effectively and accurately discover the valuable information that is hidden in the data. Many methods have been designed for the count data of microbiome. However, to our knowledge, there are only a few methods developed for the continuous microbiome data. Many microbiome data have an over-dispersed and zero-inflated data structure. Traditional methods rarely characterize this data structure and only focus on the differences in the abundance between different samples. In this study, we introduce a novel method, the zero-inflated gamma (ZIG) omnibus test, to specifically test the continuous and zero-inflated microbiome data. In this test, abundance will be tested along with zero prevalence and dispersion. We compared this method with five other popular methods. We found that ZIG omnibus test has significantly higher power and a similar or lower false-positive rate than the competing methods in the tests of simulated data. It also found more proved microbiomes in the real data with tonsil cancer. So, we conclude that ZIG omnibus test is a robust method across various biological conditions in the differential expression test of microbiome data.

The ability of predicting future frames in video sequences, known as video prediction, is an appealing yet challenging task in computer vision. This task requires an in-depth representation of video sequences and a deep understanding of real-word causal rules. Existing approaches for tackling the video prediction problem can be classified into two categories: deterministic and stochastic methods. Deterministic methods lack the ability of generating possible future frames and often yield blurry predictions. On the other hand, although current stochastic approaches can predict possible future frames, their models lack the ability of action control in the sense that they cannot generate the desired future frames conditioned on a specific action. In this paper, we propose new generative adversarial networks (GANs) for stochastic video prediction. Our framework, called VPGAN, employs an adversarial inference model and a cycle-consistency loss function to empower the framework to obtain more accurate predictions. In addition, we incorporate a conformal mapping network structure into VPGAN to enable action control for generating desirable future frames. In this way, VPGAN is able to produce fake videos of an object moving along a specific direction. Experimental results show that the combination of VPGAN with a pre-trained image segmentation model outperforms existing stochastic video prediction methods.

Background Feature selection of multi-omics data analysis remains challenging owing to the size of omics datasets, comprising approximately $$10^2$$ 10 2 –$$10^5$$ 10 5 features. In particular, appropriate methods to weight individual omics datasets are unclear, and the approach adopted has substantial consequences for feature selection. In this study, we extended a recently proposed kernel tensor decomposition (KTD)-based unsupervised feature extraction (FE) method to integrate multi-omics datasets obtained from common samples in a weight-free manner. Method KTD-based unsupervised FE was reformatted as the collection of kernelized tensors sharing common samples, which was applied to synthetic and real datasets. Results The proposed advanced KTD-based unsupervised FE method showed comparative performance to that of the previously proposed KTD method, as well as tensor decomposition-based unsupervised FE, but required reduced memory and central processing unit time. Moreover, this advanced KTD method, specifically designed for multi-omics analysis, attributes P values to features, which is rare for existing multi-omics–oriented methods. Conclusions The sample R code is available at https://github.com/tagtag/MultiR/.

Analysis of single-cell multiomics datasets is a novel topic and is considerably challenging because such datasets contain a large number of features with numerous missing values. In this study, we implemented a recently proposed tensor-decomposition (TD)-based unsupervised feature extraction (FE) technique to address this difficult problem. The technique can successfully integrate single-cell multiomics data composed of gene expression, DNA methylation, and accessibility. Although the last two have large dimensions, as many as ten million, containing only a few percentage of nonzero values, TD-based unsupervised FE can integrate three omics datasets without filling in missing values. Together with UMAP, which is used frequently when embedding single-cell measurements into two-dimensional space, TD-based unsupervised FE can produce two-dimensional embedding coincident with classification when integrating single-cell omics datasets. Genes selected based on TD-based unsupervised FE are also significantly related to reasonable biological roles.

Analysis of single-cell multiomics datasets is a novel topic and is considerably challenging because such datasets contain a large number of features with numerous missing values. In this study, we implemented a recently proposed tensor-decomposition (TD)–based unsupervised feature extraction (FE) technique to address this difficult problem. The technique can successfully integrate single-cell multiomics data composed of gene expression, DNA methylation, and accessibility. Although the last two have large dimensions, as many as ten million, containing only a few percentages of non-zero values, TD-based unsupervised FE can integrate three omics datasets without filling missing values. Together with UMAP, which is used frequently when embedding single-cell measurements into two-dimensional space, TD-based unsupervised FE can produce two-dimensional embedding coincident with classification when integrating single-cell omics datasets. Genes selected based on TD-based unsupervised FE were also significantly related to reasonable biological roles.

The prediction of drug candidates for given tissues of organisms based on expression data is a critical biological problem. By correctly predicting drug candidates for given tissues, biologists can (1) avoid an experimental process of high-throughput screening that requires excessive time and costly equipment and (2) accelerate the drug discovery process by automatically assigning drug candidates. Although high throughput screening for therapeutic compounds lead to the generation of expression data, the process of correctly assigning candidate drugs based on such data remains a rigorous task. Hence, the design of high-performance machine learning (ML) algorithms is crucial for data analysts who work with clinicians. Clinicians incorporate advanced ML tools into expert and intelligent systems to improve the drug discovery process by accurately identifying drug candidates. The transfer learning approaches that are necessary to improve the prediction performance of several tasks that are involved in identifying drug candidates are presented in this paper. The performances of machine learning algorithms are compared in the transfer learning setting by employing several evaluation measures on real data that are obtained from experiments conducted on rats to identify drug candidates. The experimental results show that the proposed transfer learning approaches outperform baseline approaches in terms of prediction performance and statistical significance.

Predicting the response, or sensitivity, of a clinical drug to a specific cancer type is an important research problem. By predicting the clinical drug response correctly, clinicians are able to understand patient-to-patient differences in drug sensitivity outcomes, which in turn results in lesser time spent and lower cost associated with identifying effective drug candidates. Although technological advances in high-throughput drug screening in cells led to the generation of a substantial amount of relevant data, the analysis of such data would be a challenging task. There is a critical need for advanced machine learning (ML) algorithms to generate accurate predictions of clinical drug response. A major goal of this work is to provide advanced ML tools to data analysts, who would in turn build prediction calculators to be incorporated into intelligent clinical decision support systems. Such innovative tools could be used to enhance patient-care, among other uses. To achieve this goal, we develop new ML techniques, including a transfer learning approach coupled with or without a boosting technique. Experimental results on real clinical data pertaining to breast cancer, multiple myeloma, and triple-negative cancer patients demonstrate the effectiveness and superiority of the proposed approaches compared to baseline approaches, including existing transfer learning methods.

The large p small n problem is a challenge without a de facto standard method available to it. In this study, we propose a tensor-decomposition (TD)-based unsupervised feature extraction (FE) formalism applied to multiomics datasets, in which the number of features is more than 100,000 whereas the number of samples is as small as about 100, hence constituting a typical large p small n problem. The proposed TD-based unsupervised FE outperformed other conventional supervised feature selection methods, random forest, categorical regression (also known as analysis of variance, or ANOVA), penalized linear discriminant analysis, and two unsupervised methods, multiple non-negative matrix factorization and principal component analysis (PCA) based unsupervised FE when applied to synthetic datasets and four methods other than PCA based unsupervised FE when applied to multiomics datasets. The genes selected by TD-based unsupervised FE were enriched in genes known to be related to tissues and transcription factors measured. TD-based unsupervised FE was demonstrated to be not only the superior feature selection method but also the method that can select biologically reliable genes. To our knowledge, this is the first study in which TD-based unsupervised FE has been successfully applied to the integration of this variety of multiomics measurements.

The accurate prediction of new interactions between drugs is important for avoiding unknown (mild or severe) adverse reactions to drug combinations. The development of effective in silico methods for evaluating drug interactions based on gene expression data requires an understanding of how various drugs alter gene expression. Current computational methods for the prediction of drug-drug interactions (DDIs) utilize data for known DDIs to predict unknown interactions. However, these methods are limited in the absence of known predictive DDIs. To improve DDIs’ interpretation, a recent study has demonstrated strong non-linear (i.e., dosedependent) effects of DDIs. In this study, we present a new unsupervised learning approach involving tensor decomposition (TD)-based unsupervised feature extraction (FE) in 3D. We utilize our approach to reanalyze available gene expression profiles for Saccharomyces cerevisiae. We found that non-linearity is possible, even for single drugs. Thus, non-linear dose-dependence cannot always be attributed to DDIs. Our analysis provides a basis for the design of effective methods for evaluating DDIs.

Analyzing single-cell pancreatic data would play an important role in understanding various metabolic diseases and health conditions. Due to the sparsity and noise present in such single-cell gene expression data, analyzing various functions related to the inference of gene regulatory networks, derived from single-cell data, remains difficult, thereby posing a barrier to the deepening of understanding of cellular metabolism. Since recent studies have led to the reliable inference of single-cell gene regulatory networks (SCGRNs), the challenge of discriminating between SCGRNs has now arisen. By accurately discriminating between SCGRNs (e.g., distinguishing SCGRNs of healthy pancreas from those of T2D pancreas), biologists would be able to annotate, organize, visualize, and identify common patterns of SCGRNs for metabolic diseases. Such annotated SCGRNs could play an important role in speeding up the process of building large data repositories. In this study, we aimed to contribute to the development of a novel deep learning (DL) application. First, we generated a dataset consisting of 224 SCGRNs belonging to both T2D and healthy pancreas and made it freely available. Next, we chose seven DL architectures, including VGG16, VGG19, Xception, ResNet50, ResNet101, DenseNet121, and DenseNet169, trained each of them on the dataset, and checked prediction based on a test set. We evaluated the DL architectures on an HP workstation platform with a single NVIDIA GeForce RTX 2080Ti GPU. Experimental results on the whole dataset, using several performance measures, demonstrated the superiority of VGG19 DL model in the automatic classification of SCGRNs, derived from the single-cell pancreatic data.

Single-cell gene regulatory network (SCGRN) inference refers to the process of inferring gene regulatory networks from single-cell data, which are generated via single-cell RNA-sequencing (scRNA-seq) technologies. Although scRNA-seq leads to the generation of data pertaining to cells of particular interest, the single-cell data are noisy and highly sparse, which makes the analysis of such data a challenging task. In this study, we model an SCGRN as a directed graph where an edge from a source node (also called transcription factor (TF)) to a target node (also called target gene) indicates that a TF regulates a target gene. Inferring the SCGRN via predicting TF-target gene regulations would help biologists better understand various diseases in terms of networks. Following the modeling step, we propose three machine learning approaches. The first approach considers feature vectors encoding regulatory relationships of expressed TFs-target genes as input. The resulting model is then used to predict unseen TF-target gene regulations. The second machine learning approach constructs new feature vectors via incorporating features obtained from stacked autoencoders, which are provided to a machine learning algorithm to induce a model and predict unseen regulations of TFs-target genes. The third approach extends the second approach via including topological features extracted from an SCGRN. We perform an experimental study comparing our approaches against adapted unsupervised approaches. Experimental results on SCGRNs pertaining to healthy and type 2 pancreatic diabetes demonstrate the clinical importance and the accurate prediction performance of the proposed approaches.

Cancer is a complex disease that is caused by rapid alteration of genes. Prediction of the state of cancer in advance contributes to a better understanding of its mechanism and improves the cancer therapy process. For example, predicting the malignancy of tumors in advance can prevent the development of cancer through the early treatment and clinical management of tumor progression. Despite generation of extensive clinical data obtained from the high-throughput technologies, it is necessary to develop machine learning algorithms to guide the prediction process. In the study, we utilize boosting and develop three computational methods to increase the performance of support vector machines (SVM). The aforementioned methods improve the performance over existing state-of-the-art algorithms, including SVM and xgboost. We evaluate the proposed boosting approach relative to the existing algorithms by using several gene expression data related to oral cancer, breast cancer, pheochromocytomas and paragangliomas, bladder cancer, and gastric cancer. The reported results using several performance measures indicate that algorithms employing the proposed approach outperform algorithms employing the baseline approach.

Background subtraction, or foreground detection, is a challenging problem in video processing. This problem is mainly concerned with a binary classification task, which designates each pixel in a video sequence as belonging to either the background or foreground scene. Traditional approaches for tackling this problem lack the power of capturing deep information in videos from a dynamic environment encountered in real-world applications, thus often achieving low accuracy and unsatisfactory performance. In this paper, we introduce a new 3-D atrous convolutional neural network, used as a deep visual feature extractor, and stack convolutional long short-term memory (ConvLSTM) networks on top of the feature extractor to capture long-term dependences in video data. This novel architecture is named a 3-D atrous ConvLSTM network. The new network can capture not only deep spatial information but also long-term temporal information in the video data. We train the proposed 3-D atrous ConvLSTM network with focal loss to tackle the class imbalance problem commonly seen in background subtraction. Experimental results on a wide range of videos demonstrate the effectiveness of our approach and its superiority over existing methods.

Traditional machine learning approaches to drug sensitivity prediction assume that training data and test data must be in the same feature space and have the same underlying distribution. However, in real-world applications, this assumption does not hold. For example, we sometimes have limited training data for the task of drug sensitivity prediction in multiple myeloma patients (target task), but we have sufficient auxiliary data for the task of drug sensitivity prediction in patients with another cancer type (related task), where the auxiliary data for the related task are in a different feature space or have a different distribution. In such cases, transfer learning, if applied correctly, would improve the performance of prediction algorithms on the test data of the target task via leveraging the auxiliary data from the related task. In this paper, we present two transfer learning approaches that combine the auxiliary data from the related task with the training data of the target task to improve the prediction performance on the test data of the target task. We evaluate the performance of our transfer learning approaches exploiting three auxiliary data sets and compare them against baseline approaches using the area under the receiver operating characteristic curve on the test data of the target task. Experimental results demonstrate the good performance of our approaches and their superiority over the baseline approaches when auxiliary data are incorporated.

Although single cell RNA sequencing (scRNA-seq) technology is newly invented and promising one, because of lack of enough information that labels individual cells, it is hard to interpret the obtained gene expression of each cell. Because of this insufficient information available, unsupervised clustering, e.g., t-Distributed Stochastic Neighbor Embedding and Uniform Manifold Approximation and Projection, is usually employed to obtain low dimensional embedding that can help to understand cell-cell relationship. One possible drawback of this strategy is that the outcome is highly dependent upon genes selected for the usage of clustering. In order to fulfill this requirement, there are many methods that performed unsupervised gene selection. In this study, a tensor decomposition (TD) based unsupervised feature extraction (FE) was applied to the integration of two scRNA-seq expression profiles that measure human and mouse midbrain development. TD based unsupervised FE could not only select coincident genes between human and mouse, but also biologically reliable genes. Coincidence between two species as well as biological reliability of selected genes is increased compared with principal component analysis (PCA) based FE applied to the same data set in the previous study. Since PCA based unsupervised FE outperformed other three popular unsupervised gene selection methods, highly variable genes, bimodal genes and dpFeature, TD based unsupervised FE can do so as well. In addition to this, ten transcription factors (TFs) that might regulate selected genes and might contribute to midbrain development are identified. These ten TFs, BHLHE40, EGR1, GABPA, IRF3, PPARG, REST, RFX5, STAT3, TCF7L2, and ZBTB33, were previously reported to be related to brain functions and diseases. TD based unsupervised FE is a promising method to integrate two scRNA-seq profiles effectively.

In this work, we extended the recently developed tensor decomposition (TD) based unsupervised feature extraction (FE) to a kernel-based method through a mathematical formulation. Subsequently, the kernel TD (KTD) based unsupervised FE was applied to two synthetic examples as well as real data sets, and the findings were compared with those obtained previously using the TD-based unsupervised FE approaches. The KTD-based unsupervised FE outperformed or performed comparably with the TD-based unsupervised FE in large p small n situations, which are situations involving a limited number of samples with many variables (observations). Nevertheless, the KTD-based unsupervised FE outperformed the TD-based unsupervised FE in non large p small n situations. In general, although the use of the kernel trick can help the TD-based unsupervised FE gain more variations, a wider range of problems may also be encountered. Considering the outperformance or comparable performance of the KTD-based unsupervised FE compared to the TD-based unsupervised FE when applied to large p small n problems, it is expected that the KTD-based unsupervised FE can be applied in the genomic science domain, which involves many large p small n problems, and, in which, the TD-based unsupervised FE approach has been effectively applied.

In this paper we present a new approach, named DLGraph, for malware detection using deep learning and graph embedding. DLGraph employs two stacked denoising autoencoders (SDAs) for representation learning, taking into consideration computer programs' function-call graphs and Windows application programming interface (API) calls. Given a program, we first use a graph embedding technique that maps the program's function-call graph to a vector in a low-dimensional feature space. One SDA in our deep learning model is used to learn a latent representation of the embedded vector of the function-call graph. The other SDA in our model is used to learn a latent representation of the given program's Windows API calls. The two learned latent representations are then merged to form a combined feature vector. Finally, we use softmax regression to classify the combined feature vector for predicting whether the given program is malware or not. Experimental results based on different datasets demonstrate the effectiveness of the proposed approach and its superiority over a related method.

Transfer learning (TL) algorithms aim to improve the prediction performance in a target task (e.g. the prediction of cisplatin sensitivity in triple-negative breast cancer patients) via transferring knowledge from auxiliary data of a related task (e.g. the prediction of docetaxel sensitivity in breast cancer patients), where the distribution and even the feature space of the data pertaining to the tasks can be different. In real-world applications, we sometimes have a limited training set in a target task while we have auxiliary data from a related task. To obtain a better prediction performance in the target task, supervised learning requires a sufficiently large training set in the target task to perform well in predicting future test examples of the target task. In this paper, we propose a TL approach for cancer drug sensitivity prediction, where our approach combines three techniques. First, we shift the representation of a subset of examples from auxiliary data of a related task to a representation closer to a target training set of a target task. Second, we align the shifted representation of the selected examples of the auxiliary data to the target training set to obtain examples with representation aligned to the target training set. Third, we train machine learning algorithms using both the target training set and the aligned examples. We evaluate the performance of our approach against baseline approaches using the Area Under the receiver operating characteristic (ROC) Curve (AUC) on real clinical trial datasets pertaining to multiple myeloma, nonsmall cell lung cancer, triple-negative breast cancer, and breast cancer. Experimental results show that our approach is better than the baseline approaches in terms of performance and statistical significance.

Predicting cancer disease state is an important problem in the cancer discovery process. For example, discriminating between benign and malignant tumors improves the medical diagnosis of cancer. Although technological advancing led to the generation of data pertaining to patients with different disease states, evaluating the prediction performance of machine learning algorithms would be an important step. In this paper, we propose using machine learning algorithms such as a variant of AdaBoost, deepboost, xgboost and support vector machines and evaluate them using area under curve and accuracy on real clinical data related to thyroid cancer, colon cancer and liver cancer. Experimental results show the good performance of SVM.

Reverse engineering gene regulatory networks (GRNs), also known as GRN inference, refers to the process of reconstructing GRNs from gene expression data. A GRN is modeled as a directed graph in which nodes represent genes and edges show regulatory relationships between the genes. By predicting the edges to infer a GRN, biologists can gain a better understanding of regulatory circuits and functional elements in cells. Many bioinformatics tools have been developed to computationally reverse engineer GRNs. However, none of these tools is able to perform perfect GRN inference. In this paper, we propose a graph mining approach capable of discovering frequent patterns from the GRNs inferred by existing methods. These frequent or common patterns are more likely to occur in true regulatory networks. Experimental results on different datasets demonstrate the good quality of the discovered patterns, and the superiority of our approach over the existing GRN inference methods.

In systems biology, two important regulatory networks are gene regulatory networks (GRNs) and regulatory networks of microRNAs (RNMs). A GRN is modeled as a directed graph in which a node represents a gene or transcription factor (TF), and an edge from a TF to a gene indicates that the TF regulates the expression of the gene. An RNM is modeled as a bipartite directed graph with two disjoint sets of nodes: a set of nodes that represent microRNAs (miRNAs) and a set of nodes that represent genes or TFs. Directed edges between these two sets of nodes represent miRNA-target interactions or TF-miRNA regulatory relations. In this paper, we present an approach to reverse engineering GRNs and RNMs using a dynamic Bayesian network and mutual information scoring function. Our approach is able to automatically infer both GRNs and RNMs from time series of expression data. Experimental results on different datasets show that our approach is more accurate than other time-series based network inference methods.

Background Predicting the response to a drug for cancer disease patients based on genomic information is an important problem in modern clinical oncology. This problem occurs in part because many available drug sensitivity prediction algorithms do not consider better quality cancer cell lines and the adoption of new feature representations; both lead to the accurate prediction of drug responses. By predicting accurate drug responses to cancer, oncologists gain a more complete understanding of the effective treatments for each patient, which is a core goal in precision medicine. Results In this paper, we model cancer drug sensitivity as a link prediction, which is shown to be an effective technique. We evaluate our proposed link prediction algorithms and compare them with an existing drug sensitivity prediction approach based on clinical trial data. The experimental results based on the clinical trial data show the stability of our link prediction algorithms, which yield the highest area under the ROC curve (AUC) and are statistically significant. Conclusions We propose a link prediction approach to obtain new feature representation. Compared with an existing approach, the results show that incorporating the new feature representation to the link prediction algorithms has significantly improved the performance. Electronic supplementary material The online version of this article (doi:10.1186/s12918-017-0463-8) contains supplementary material, which is available to authorized users.

Reverse engineering gene regulatory networks (GRNs), also known as network inference, refers to the process of reconstructing GRNs from gene expression data. Biologists model a GRN as a directed graph in which nodes represent genes and links show regulatory relationships between the genes. By predicting the links to infer a GRN, biologists can gain a better understanding of regulatory circuits and functional elements in cells. Existing supervised GRN inference methods work by building a feature-based classifier from gene expression data and using the classifier to predict the links in GRNs. Observing that GRNs are sparse graphs with few links between nodes, we propose here to use under-sampling, over-sampling and boosting techniques to enhance the prediction performance. Experimental results on different datasets demonstrate the good performance of the proposed approach and its superiority over the existing methods.

Gene regulation is a series of processes that control gene expression and its extent. The connections among genes and their regulatory molecules, usually transcription factors, and a descriptive model of such connections, are known as gene regulatory networks (GRNs). Elucidating GRNs is crucial to understand the inner workings of the cell and the complexity of gene interactions. To date, numerous algorithms have been developed to infer gene regulatory networks. However, as the number of identified genes increases and the complexity of their interactions is uncovered, networks and their regulatory mechanisms become cumbersome to test. Furthermore, prodding through experimental results requires an enormous amount of computation, resulting in slow data processing. Therefore, new approaches are needed to expeditiously analyze copious amounts of experimental data resulting from cellular GRNs. To meet this need, cloud computing is promising as reported in the literature. Here we propose new MapReduce algorithms for inferring gene regulatory networks on a Hadoop cluster in a cloud environment. These algorithms employ an information-theoretic approach to infer GRNs using time-series microarray data. Experimental results show that our MapReduce program is much faster than an existing tool while achieving slightly better prediction accuracy than the existing tool., Comment: 19 pages, 5 figures

Supervised methods for inferring gene regulatory networks (GRNs) perform well with good training data. However, when training data is absent, these methods are not applicable. Unsupervised methods do not need training data but their accuracy is low. In this paper, we combine supervised and unsupervised methods to infer GRNs using time-series gene expression data. Specifically, we use results obtained from unsupervised methods to train supervised methods. Since the results contain noise, we develop a data cleaning algorithm to remove noise, hence improving the quality of the training data. These refined training data are then used to guide classifiers including support vector machines and deep learning tools to infer GRNs through link prediction. Experimental results on several data sets demonstrate the good performance of the classifiers and the effectiveness of our data cleaning algorithm.

Predicting mortality of Middle East respiratory syndrome (MERS) patients with identified outcomes is a core goal for hospitals in deciding whether a new patient should be hospitalized or not in the presence of limited resources of the hospitals. We present an oversampling approach that we call Greedy-Based Oversampling Approach (GBOA). We evaluate our approach and compare it against the standard oversampling approach from a classification perspective on real dataset collected from the Saudi Ministry of Health using two popular supervised classification methods, Random Forests and Support Vector Machines. Our results demonstrate that our approach outperforms the other standard approach from a classification perspective by giving the highest accuracy with statistical significance on the 20 simulations of the real dataset.

Network inference through link prediction is an important data mining problem that finds many applications in computational social science and biomedicine. For example, by predicting links, i.e., regulatory relationships, between genes to infer gene regulatory networks (GRNs), computational biologists gain a better understanding of the functional elements and regulatory circuits in cells. Unsupervised methods have been widely used to infer GRNs; however, these methods often create missing and spurious links. In this paper, we propose a learning framework to improve the unsupervised methods. Given a network constructed by an unsupervised method, the proposed framework employs a graph sparsification technique for network sampling and principal component analysis for feature selection to obtain better quality training data, which guides three classifiers to predict and clean the links of the given network. The three classifiers include neural networks, random forests and support vector machines. Experimental results on several datasets demonstrate the good performance of the proposed learning framework and the classifiers used in the framework.

Link prediction is an important data mining problem that has many applications in different domains such as social network analysis and computational biology. For example, biologists model gene regulatory networks (GRNs) as directed graphs where nodes are genes and links show regulatory relationships between the genes. By predicting links in GRNs, biologists can gain a better understanding of the cell regulatory circuits and functional elements. Existing supervised methods for GRN inference work by building a feature-based classifier from gene expression data and using the classifier to predict links in the GRNs. In this paper we present a new supervised approach for link prediction in GRNs. Our approach employs both gene expression data and topological features extracted from the GRNs, in combination with three machine learning algorithms including random forests, support vector machines and neural networks. Experimental results on different datasets demonstrate the good performance of the proposed approach and its superiority over the existing methods.

Background Programs based on hash tables and Burrows-Wheeler are very fast for mapping short reads to genomes but have low accuracy in the presence of mismatches and gaps. Such reads can be aligned accurately with the Smith-Waterman algorithm but it can take hours and days to map millions of reads even for bacteria genomes. Results We introduce a GPU program called MaxSSmap with the aim of achieving comparable accuracy to Smith-Waterman but with faster runtimes. Similar to most programs MaxSSmap identifies a local region of the genome followed by exact alignment. Instead of using hash tables or Burrows-Wheeler in the first part, MaxSSmap calculates maximum scoring subsequence score between the read and disjoint fragments of the genome in parallel on a GPU and selects the highest scoring fragment for exact alignment. We evaluate MaxSSmap’s accuracy and runtime when mapping simulated Illumina E.coli and human chromosome one reads of different lengths and 10% to 30% mismatches with gaps to the E.coli genome and human chromosome one. We also demonstrate applications on real data by mapping ancient horse DNA reads to modern genomes and unmapped paired reads from NA12878 in 1000 genomes. Conclusions We show that MaxSSmap attains comparable high accuracy and low error to fast Smith-Waterman programs yet has much lower runtimes. We show that MaxSSmap can map reads rejected by BWA and NextGenMap with high accuracy and low error much faster than if Smith-Waterman were used. On short read lengths of 36 and 51 both MaxSSmap and Smith-Waterman have lower accuracy compared to at higher lengths. On real data MaxSSmap produces many alignments with high score and mapping quality that are not given by NextGenMap and BWA. The MaxSSmap source code in CUDA and OpenCL is freely available from http://www.cs.njit.edu/usman/MaxSSmap. Electronic supplementary material The online version of this article (doi:10.1186/1471-2164-15-969) contains supplementary material, which is available to authorized users.

Ensemble methods such as AdaBoost are popular machine learning methods that create highly accurate classifier by combining the predictions from several classifiers. We present a parametrized method of AdaBoost that we call Top-k Parametrized Boost. We evaluate our and other popular ensemble methods from a classification perspective on several real datasets. Our empirical study shows that our method gives the minimum average error with statistical significance on the datasets.