The detection of miRNA regulatory modules (MRMs) can facilitate the analysis of miRNA combination effects in aberrant transcriptional regulatory networks. Existing methods suffer from stochastic or require a predetermined number of regulatory modules. We here develop a parameter-free computational framework named DeMine to predict MRMs. Briefly, DeMine is an information entropy-based method implemented by three steps. It first transforms miRNA regulatory network into a miRNA-miRNA synergistic network, and then detects miRNA clusters by maximizing the referred cluster entropy density. After that, the co-regulated mRNAs are added into corresponding clusters to form final MRMs. Compared with existing methods Mirsynergy and BCM on synthetic dataset and three real cancer datasets: ovarian cancer (OVCA), breast cancer (BRCA) and thyroid cancer (THCA) datasets, the proposed method can not only exhibit higher accuracy but also extract larger cohesiveness-preserved MRMs. More importantly, DeMine can find more miRNAs as tumor markers for the diagnosis of tumors.
In this paper, I propose two novel methods for extracting synchronously fluctuated genes (SFGs) from a transcriptome data. Variability and synchrony in biological signals are generally considered to be associated with the system's stability in some sense. However, a standard method for extracting SFGs from a transcriptome data with high reproducibility has not been established. Here, I propose two novel methods for extracting SFGs. The first method has two steps: selection of remarkably fluctuated genes and extraction of synchronized gene clusters. The other method is based on principal component analysis. It has been confirmed that the two methods have high extraction performance for artificial data and a moderate level of reproducibility for real data. The proposed methods will help to extract candidate genes related to the stability and homeostasis in living organisms.
Bacterial whole-genome sequences have recently become widely available via innovative and rapid progress in technologies such as high-throughput sequencing and computing. Genomes of environmental microorganisms have also been sequenced, and their number is expected to increase in the future. Typically, phylogenetic analysis is performed after genome sequencing of such organisms. 16S rRNA is a standard locus for the phylogenetic analysis of prokaryotes. However, 16S rRNA phylogenetic trees are not always reliable because of out-paralogs and horizontal gene transfer. To overcome this problem, multiple genes (or proteins) should be employed. Therefore, we developed “Genome Identifier, ” which can be used for constructing a concatenated phylogenetic tree in the form of a species tree by predicting genes from newly sequenced genomic data and collecting homologous sequences from other species.
Conventional computerbased methods for identifying antifungal peptides (AFPs) are primarily based on large amounts of task-specific knowledge in feature extraction. This paper introduces a method, called AFPDeep, which only uses peptide sequences as input, without calculating any features by human intervention for the prediction of AFPs. First, an embedding layer is used to automatically code the sequence and to learn a dense vector representation for each kind of amino acid appearing in the training dataset. Second, a convolutional neural network (CNN) layer followed by a long short term memory (LSTM) layer is used to capture the local clues that are good indicators of AFPs, and to learn the long-term dependence and contextual information of the input data in a flexible way. Lastly, all the above layers are intertwined into hybrid neural network architecture for antifungal peptide prediction. Upon comparison with other methods on the same datasets, the AFPDeep yielded competitive results, achieving an AUC of on the Antifp_Main dataset, on the Antifp_DS1 dataset, and on the Antifp_DS2 dataset. These results suggest that the hybrid architecture of CNN-LSTM combined with character embedding is an effective model for improving the prediction of AFPs.