Search

Deep learning methods have been successfully used in a variety of different contexts and achieved state of the art performance in many different tasks. In this paper, we explore the performance of deep learning methods in the task of predicting functional genetic variant. First, we test the performance of a few types of neural network models in making prediction using only DNA sequence. The result shows that convolutional neural network (CNN) has the best performance. Second, we explore the possibility of forming a hybrid network to make prediction with both DNA sequence and evolutionary nucleotide conservation information as input. We observe a better performance than using only conservation information by applying a dropout mask for the transformed feature of DNA sequence. We further discuss this technique as a possible common solution for combining features of different powers.

Maize is one of the most important crops worldwide. To understand the biological processes underlying various traits of the crop, e.g. yield and response to stress, a detailed protein-protein interaction (PPI) network is highly demanded. Unfortunately, there are very few such PPIs available in literature. Therefore, in this work, we present the PPIM (Protein-Protein Interaction Database for Maize) database [1], which covers 2,762,560 interactions among 14,000 proteins. The PPIM databases contains not only accurately predicted PPIs but also those molecular interactions collected from literature. The database is freely available at http://comp-sysbio.org/ppim with a user-friendly powerful interface. We believe the PPIM resource can help biologists better understand the maize crop.

Although human and macaque share a lot of genes, their brains have different functions. In this study, we compare human and macaque brains from a systematic perspective by assuming that the difference between human and macaque brains is due to the difference of molecular circuits underlying brains instead of certain individual genes. Specially, we construct a gene interaction network for each brain region and investigate the difference between human and macaque brains based on those region-specific gene networks. We found that the genes in human brain networks tend to interact with each other compared with those of macaque brain networks. In addition, the human genes are more likely to form functional modules compared against macaque genes. Moreover, the human specific genes tend to interact with hub genes, while both the hub genes and human specific genes are enriched in brain functions. These findings can help better understand the functional divergence between human and macaque brains.

The timely and accurate diagnosis of Alzheimer's Disease (AD) is important for preventing the progress of the irreversible disease. Recently, various types of imaging techniques, e.g. Magnetic Resonance Imaging (MRI) and Positron Emission Tomography (PET), have been widely used for the diagnosis of AD. Despite their usefulness, the image data generated are high-dimensional and noisy which make it difficult to give accurate diagnosis. In this paper, we propose a novel feature selection approach to detect the informative features from MRI data, which improves both the diagnosis accuracy and reduces the computational cost. Benchmarking results on the AD challenge datasets show that our proposed approach outperforms other popular feature selection methods. Furthermore, compared with the top winners of the challenge, our approach performs best in diagnosis of AD and has comparable performance with others in the detection of mild cognitive impairments (MCIs). The good performance on the real data demonstrate that our proposed approach is effective in prediction of ADs.

Phosphorylation is a post-translational modification process mediated by kinases through the addition of a covalently bound phosphate group, which plays important roles in a wide range of cellular progresses, such as signaling cascades and development. Over the past years, despite many phosphorylation sites have been determined with mass spectrometry techniques, it is not clear which kinase phosphorylates which proteins. Under the circumstance, we propose a new probabilistic model to identify the substrates phosphorylated by certain kinases. Furthermore, we construct three tissue-specific phosphorylation networks based on protein expression data. Investigating the constructed tissue-specific networks, we find they are functionally consistent with the corresponding tissues, implying the effectiveness and biological significance of our proposed approach.

With the development of industrial technology, human beings are suffering from the pollution of heavy metals. Arsenic is a trace element of toxin, which has become a potential threat to rice that assimilates arsenic. Identifying the signaling pathways induced by arsenate stress in rice can help us understand how arsenic affects the biological systems of rice. In this article, based on rice molecular interaction network, we identified the signaling pathways induced by arsenic in rice with a computational approach. Functional enrichment analysis indicates that the predicted signaling pathway is related to arsenate stress in rice.

The identification of drug-target interactions is a crucial step in the drug-discovery process. It has been suggested that drug-target interactions are driven by drug-domain interactions. Based on the integration of two recently published datasets, i.e., Drug-target interactions in myocardial infarction (My-DTome) and drug-domain interaction network, this paper reports the association between drugs and protein domains in the context of myocardial infarction (MI). A MI drug-domain interaction network, My-DDome, was constructed. The functional similarity between domains based on their Gene Ontology (GO) annotations was estimated. The association between domains and therapeutic effects was investigated. Lists of GO annotations and Anatomical Therapeutic Chemical classification (ATC) codes highly enriched in My-DDome were identified. We show that drugs acting on blood and blood forming organs (ATC code B) and sensory organs (ATC code S) are significantly enriched in My-DDome (p < 0.000001). Top enriched GO terms include GO:0003824 (catalytic activity), GO:0008152 (metabolic process) and GO:0030170 (pyridoxal phosphate binding). By incorporating protein domain information into My-DTome, more detailed insights into the interplay between drugs, their known targets and seemingly unrelated proteins are provided.

Prostate cancer is one of the most important public health problems in developed countries. To date, a systematic understanding of the pathogenesis of prostate cancer is still lacking. In this work, we identified differentially expressed protein-coding genes and long non-coding RNAs (lncRNAs) between normal and cancer tissues based on a recent RNA-seq study from Caucasian population. We then investigated the relationship between differentially expressed genes and lncRNAs. Furthermore, based on a recently published prostate cancer study on Chinese population, we identified differentially expressed genes between Caucasian and Chinese populations to investigate racial difference. Moreover, for the first time, we compared the correlation of lncRNA-gene across populations. In the end, a lot of differentially expressed genes and lncRNAs were identified. Our results revealed that most of the lncRNA-gene pairs were positively correlated especially for the lncRNA-host gene pairs, indicating the probable mechanism of lncRNA. And 320 genes were differentially expressed in prostate cancer across populations, which may help us to investigate the ethnic differences of prostate cancer. In addition, our results suggested that lncRNAs regulate genes in different manners across populations. Our findings may help understand molecular events underlying prostate cancer development.

Identifying dysregulation modules for complex diseases, such as B-cell lymphomas, can provide insights into the mechanisms of diseases and help to identify novel drug targets. In this work, based on molecular interaction network, we applied a network flow model to identify the dysregulation modules for three subtypes of non-Hodgkin's lymphomas, including Burkitt's lymphoma (BL), follicular lymphoma (FL), and mantle cell lymphoma (MCL). In our identified dysregulation modules, there are multiple genes that were reported in literature to be related to B-cell lymphomas, which demonstrate that our presented method is really effective for identifying dysregulation modules related to diseases.

Protein subcellular localization provides insights into protein function. In literature, various computational methods have been developed for this problem based on protein sequences, where most methods have limited prediction accuracy. Therefore, a general computational method with high prediction accuracy is necessary. In this work, we present a novel balanced ensemble classifier for fungi protein subcellular localization prediction based only on protein sequences. We make three fold contributions to this filed. First, we present a new algorithm to cope with imbalance problem that arises in protein subcellular localization prediction, which can improve prediction accuracy significantly. Second, we employ feature selection techniques to find out most informative features for each compartment, and reduce computation cost and improve prediction accuracy at the same time. Third, an ensemble classifier combing outputs from distinct classifiers is presented to further improve prediction accuracy. The numerical results on benchmark dataset demonstrate the efficiency and effectiveness of the proposed method.

Microarray technology is a useful tool for monitoring the expressed levels of thousands of genes simultaneously. Recently, mixture modelling has been used to extract information from expressed genes. It utilizes two separate steps to estimate the number of classes and model parameters, respectively, which however may be time-consuming and fall into sub-optimal solutions. In this paper, we therefore apply an one-step approach, namely rival penalized expectation-maximization (RPEM) algorithm, to microarray data analysis. The RPEM algorithm is capable of estimating the parameters of the normal mixture model, meanwhile determining the number of classes automatically. The numerical results have shown the effectiveness of this technique on real microarray data analysis.

In this paper, we propose a new representation of DNA sequences, which constructs the word frequency vector with multiple resolutions based on the chaos game representation. Compared with the traditional vector, it combines a range of resolutions and reserves higher resolutions, but the dimension is reduced greatly relatively. The algorithm is detailed, which calculates coding format and codes each sequence. To evaluate the significance of our method, we represent Alu sequences by our proposed coding format. After that, the acquired vectors are used to train BP neural networks to recognize the Alu sequences. The experimental results show that this representation of DNA sequences is significant and efficient in biological data processing.

This work presents a SVM and feature/spl I.bar/core-based algorithm for identification of key genome features. The significant difficulty in selecting key features from a high dimensional space of features is the curse of dimensions in searching. For this reason, a feature/spl I.bar/core-based search strategy is proposed in this algorithm. The strategy integrates the forward selection and backward elimination techniques. In this algorithm all key genome features are formed through the agglomeration and expansion of the feature core based on potential information about relevance in a SVM classifier. The application given proves that the algorithm is faster and more efficient than other methods such as clustering, and the single SVM-based method.