Journal of Bioinformatics and Computational Biology

Current high-throughput experimental techniques make it feasible to infer gene regulatory interactions at the whole-genome level with reasonably good accuracy. Such experimentally inferred regulatory networks have become available for a number of simpler model organisms such as S. cerevisiae, and others. The availability of such networks provides an opportunity to compare gene regulatory processes at the whole genome level, and in particular, to assess similarity of regulatory interactions for homologous gene pairs either from the same or from different species.

We present here a new technique for analyzing the regulatory interaction neighborhoods of paralogous gene pairs. Our central focus is the analysis of S. cerevisiae gene interaction graphs, which are of particular interest due to the ancestral whole-genome duplication (WGD) that allows to distinguish between paralogous transcription factors that are traceable to this duplication event and other paralogues. Similar analysis is also applied to E. coli and C. elegans networks.

We compare paralogous gene pairs according to the presence and size of bi-fan arrays, classically associated in the literature with gene duplication, within other network motifs. We further extend this framework beyond transcription factor comparison to obtain topology-based similarity metrics based on the overlap of interaction neighborhoods applicable to most genes in a given organism.

We observe that our network divergence metrics show considerably larger similarity between paralogues, especially those traceable to WGD. This is the case for both yeast and C. elegans, but not for E. coli regulatory network. While there is no obvious cross-species link between metrics, different classes of paralogues show notable differences in interaction overlap, with traceable duplications tending toward higher overlap compared to genes with shared protein families. Our findings indicate that divergence in paralogous interaction networks reflects a shared genetic origin, and that our approach may be useful for investigating structural similarity in the interaction networks of paralogous genes.

In this work, we study period control of the mammalian cell cycle via coupling with the cellular clock. For this, we make use of the oscillators’ synchronization dynamics and investigate methods of slowing down the cell cycle with the use of clock inputs. Clock control of the cell cycle is well established via identified molecular mechanisms, such as the CLOCK:BMAL1-mediated induction of the wee1 gene, resulting in the WEE1 kinase that represses the active form of mitosis promoting factor (MPF), the essential cell cycle component. To investigate the coupling dynamics of these systems, we use previously developed models of the clock and cell cycle oscillators and center our studies on unidirectional clock $\to$ cell cycle coupling. Moreover, we propose an hypothesis of a Growth Factor (GF)-responsive clock, involving a pathway of the non-essential cell cycle complex cyclin D/CDK4. We observe a variety of rational ratios of clock to cell cycle period, such as: 1:1, 3:2, 4:3, and 5:4. Finally, our protocols of period control are successful in effectively slowing down the cell cycle by the use of clock modulating inputs, some of which correspond to existing drugs.

Gene regulatory network inference is one of the central problems in computational biology. We need models that integrate the variety of data available in order to use their complementarity information to overcome the issues of noisy and limited data. BENIN: Biologically Enhanced Network INference is our proposal to integrate data and infer more accurate networks. BENIN is a general framework that jointly considers different types of prior knowledge with expression datasets to improve the network inference. The method states the network inference as a feature selection problem and uses a popular penalized regression method, the Elastic net, combined with bootstrap resampling to solve it. BENIN significantly outperforms the state-of-the-art methods on the simulated data from the DREAM 4 challenge when combining genome-wide location data, knockout gene expression data, and time series expression data.

Identifying protein complexes is an important issue in computational biology, as it benefits the understanding of cellular functions and the design of drugs. In the past decades, many computational methods have been proposed by mining dense subgraphs in Protein–Protein Interaction Networks (PINs). However, the high rate of false positive/negative interactions in PINs prevents accurately detecting complexes directly from the raw PINs. In this paper, we propose a denoising approach for protein complex detection by using variational graph auto-encoder. First, we embed a PIN to vector space by a stacked graph convolutional network (GCN), then decide which interactions in the PIN are credible. If the probability of an interaction being credible is less than a threshold, we delete the interaction. In such a way, we reconstruct a reliable PIN. Following that, we detect protein complexes in the reconstructed PIN by using several typical detection methods, including CPM, Coach, DPClus, GraphEntropy, IPCA and MCODE, and compare the results with those obtained directly from the original PIN. We conduct the empirical evaluation on four yeast PPI datasets (Gavin, Krogan, DIP and Wiphi) and two human PPI datasets (Reactome and Reactomekb), against two yeast complex benchmarks (CYC2008 and MIPS) and three human complex benchmarks (REACT, REACT_uniprotkb and CORE_COMPLEX_human), respectively. Experimental results show that with the reconstructed PINs obtained by our denoising approach, complex detection performance can get obviously boosted, in most cases by over 5%, sometimes even by 200%. Furthermore, we compare our approach with two existing denoising methods (RWS and RedNemo) while varying different matching rates on separate complex distributions. Our results show that in most cases (over 2/3), the proposed approach outperforms the existing methods.

Clustering analysis of gene expression data is essential for understanding complex biological data, and is widely used in important biological applications such as the identification of cell subpopulations and disease subtypes. In commonly used methods such as hierarchical clustering (HC) and consensus clustering (CC), holistic expression profiles of all genes are often used to assess the similarity between samples for clustering. While these methods have been proven successful in identifying sample clusters in many areas, they do not provide information about which gene sets (functions) contribute most to the clustering, thus limiting the interpretability of the resulting cluster. We hypothesize that integrating prior knowledge of annotated gene sets would not only achieve satisfactory clustering performance but also, more importantly, enable potential biological interpretation of clusters. Here we report ClusterMine, an approach that identifies clusters by assessing functional similarity between samples through integrating known annotated gene sets in functional annotation databases such as Gene Ontology. In addition to the cluster membership of each sample as provided by conventional approaches, it also outputs gene sets that most likely contribute to the clustering, thus facilitating biological interpretation. We compare ClusterMine with conventional approaches on nine real-world experimental datasets that represent different application scenarios in biology. We find that ClusterMine achieves better performances and that the gene sets prioritized by our method are biologically meaningful. ClusterMine is implemented as an R package and is freely available at: www.genemine.org/clustermine.php

MicroRNA (miRNA) sponges’ regulatory mechanisms play an important role in developing human cancer. Herein, we develop a new method to explore potential miRNA sponge interactions (EPMSIs) for breast cancer. Based on some known interactions, and a matching gene expression profile, EPMSIs explored other potential miRNA sponge interactions for breast cancer. Every interaction is inferred with a value representing interaction intensity. Then, we apply a clustering algorithm called BCPlaid to potential interactions. Ten modules are identified; nine of them are closely associated with biological enrichments. When we employ a classification algorithm to separate normal and tumor samples in each module, each module demonstrates powerful classification performance. Furthermore, EPMSI illustrates a new method to explore the miRNA sponge regulatory network for breast cancer by applying its superior performance.

Membrane proteins play essential roles in modern medicine. In recent studies, some membrane proteins involved in ectodomain shedding events have been reported as the potential drug targets and biomarkers of some serious diseases. However, there are few effective tools for identifying the shedding event of membrane proteins. So, it is necessary to design an effective tool for predicting shedding event of membrane proteins. In this study, we design an end-to-end prediction model using deep neural networks with long short-term memory (LSTM) units and attention mechanism, to predict the ectodomain shedding events of membrane proteins only by sequence information. Firstly, the evolutional profiles are encoded from original sequences of these proteins by Position-Specific Iterated BLAST (PSI-BLAST) on Uniref50 database. Then, the LSTM units which contain memory cells are used to hold information from past inputs to the network and the attention mechanism is applied to detect sorting signals in proteins regardless of their position in the sequence. Finally, a fully connected dense layer and a softmax layer are used to obtain the final prediction results. Additionally, we also try to reduce overfitting of the model by using dropout, L2 regularization, and bagging ensemble learning in the model training process. In order to ensure the fairness of performance comparison, firstly we use cross validation process on training dataset obtained from an existing paper. The average accuracy and area under a receiver operating characteristic curve (AUC) of five-fold cross-validation are 81.19% and 0.835 using our proposed model, compared to 75% and 0.78 by a previously published tool, respectively. To better validate the performance of the proposed model, we also evaluate the performance of the proposed model on independent test dataset. The accuracy, sensitivity, and specificity are 83.14%, 84.08%, and 81.63% using our proposed model, compared to 70.20%, 71.97%, and 67.35% by the existing model. The experimental results validate that the proposed model can be regarded as a general tool for predicting ectodomain shedding events of membrane proteins. The pipeline of the model and prediction results can be accessed at the following URL: http://www.csbg-jlu.info/DeepSMP/.

The automatic extraction of disease named entity is a challenging research problem that has attracted attention from the biomedical text mining community. Handcrafted feature methods were employed for this task given a little success since they are limited by the scope of the expert. Lately, deep learning-based methods have been employed to solve this issue. However, most architectures used for this task take into consideration long dependencies only. The proposed method is a two-stage deep neural network model. We start by discovering local dependencies and creating high-level features from word embedding inputs using a deep convolutional neural network. Then we identify long dependencies using a bi-directional recurrent neural network. To solve the problem of unbalanced dataset given by the BMEWO tagging schema and to enforce sequence modeling, we developed a new POS-based tagging schema that subdivides the dominant class into smaller more balanced units. The proposed system was trained and tested on NCBI and achieved an $F$-score of 85.59 outperforming the current state-of-the-art methods. Our research results show the effectiveness of using both long and short dependencies. The results also illustrate the benefits of combining different word embedding techniques and the incorporation of morphological features in this task.

Cancer is a complex disease caused by the accumulation of genetic alterations during the individual’s life. Such alterations are called genetic mutations and can be divided into two groups: (1) Passenger mutations, which are not responsible for cancer and (2) Driver mutations, which are significant for cancer and responsible for its initiation and progression. Cancer cells undergo a large number of mutations, of which most are passengers, and few are drivers. The identification of driver mutations is a key point and one of the biggest challenges in Cancer Genomics. Many computational methods for such a purpose have been developed in Cancer Bioinformatics. Such computational methods are complex and are usually described in a high level of abstraction. This tutorial details some classical computational methods, from a computational perspective, with the transcription in an algorithmic format towards an easy access by researchers.