Journal of Bioinformatics and Computational Biology

Time plays an essential role in many biological systems, especially in cell cycle. Many models of biological systems rely on differential equations, but parameter identification is an obstacle to use differential frameworks. In this paper, we present a new hybrid modeling framework that extends René Thomas’ discrete modeling. The core idea is to associate with each qualitative state “celerities” allowing us to compute the time spent in each state. This hybrid framework is illustrated by building a 5-variable model of the mammalian cell cycle. Its parameters are determined by applying formal methods on the underlying discrete model and by constraining parameters using timing observations on the cell cycle. This first hybrid model presents the most important known behaviors of the cell cycle, including quiescent phase and endoreplication.

We consider the problem of sorting signed permutations by reversals, transpositions, transreversals, and block-interchanges. The problem arises in the study of species evolution via large-scale genome rearrangement operations. Recently, Hao et al. gave a 2-approximation scheme called genome sorting by bridges (GSB) for solving this problem. Their result extended and unified the results of (i) He and Chen — a 2-approximation algorithm allowing reversals, transpositions, and block-interchanges (by also allowing transversals) and (ii) Hartman and Sharan — a 1.5-approximation algorithm allowing reversals, transpositions, and transversals (by also allowing block-interchanges). The GSB result is based on introduction of three bridge structures in the breakpoint graph, the L-bridge, T-bridge, and X-bridge that models goodreversal, transposition/transreversal, and block-interchange, respectively. However, the paper by Hao et al. focused on proving the 2-approximation GSB scheme and only mention a straightforward $O(n^6)$ algorithm. In this paper, we give an $O(n^3)$ algorithm for implementing the GSB scheme. The key idea behind our faster GSB algorithm is to represent cycles in the breakpoint graph by their canonical sequences, which greatly simplifies the search for these bridge structures. We also give some comparison results (running time and computed distances) against the original GSB implementation.

Genomic islands (GIs) are clusters of functionally related genes acquired by lateral genetic transfer (LGT), and they are present in many bacterial genomes. GIs are extremely important for bacterial research, because they not only promote genome evolution but also contain genes that enhance adaption and enable antibiotic resistance. Many methods have been proposed to predict GI. But most of them rely on either annotations or comparisons with other closely related genomes. Hence these methods cannot be easily applied to new genomes. As the number of newly sequenced bacterial genomes rapidly increases, there is a need for methods to detect GI based solely on sequences of a single genome. In this paper, we propose a novel method, GI-SVM, to predict GIs given only the unannotated genome sequence. GI-SVM is based on one-class support vector machine (SVM), utilizing composition bias in terms of k-mer content. From our evaluations on three real genomes, GI-SVM can achieve higher recall compared with current methods, without much loss of precision. Besides, GI-SVM allows flexible parameter tuning to get optimal results for each genome. In short, GI-SVM provides a more sensitive method for researchers interested in a first-pass detection of GI in newly sequenced genomes.

Recently, the description of immune response by discrete models has emerged to play an important role to study the problems in the area of human immunodeficiency virus type 1 (HIV-1) infection, leading to AIDS. As infection of target immune cells by HIV-1 mainly takes place in the lymphoid tissue, cellular automata (CA) models thus represent a significant step in understanding when the infected population is dispersed. Motivated by these, the studies of the dynamics of HIV-1 infection using CA in memory have been presented to recognize how CA have been developed for HIV-1 dynamics, which issues have been studied already and which issues still are objectives in future studies.

The tumor suppressor p53 can be activated by DNA damage and exhibits undamped pulses. Recent reports have demonstrated a non-threshold mechanism for p53 dynamics. However, no related theoretical studies have been proposed. Here, we constructed a refined DNA damage repair model that incorporated both intrinsic and extrinsic DNA lesions. We proposed that the basal DNA damage may trigger significant fractions of p53 pulses. We also reproduced the heterogeneity of p53 dynamics in experiments. The number of p53 pulses showed no correlations with DNA damage. We also replicated the linear correlation between DNA damage and the probability of igniting a pulse. Our model has unraveled the heterogeneous p53 responses.

Raman spectroscopy (including surface enhanced Raman spectroscopy (SERS) and tip enhanced Raman spectroscopy (TERS)) is a highly promising experimental method for investigations of biomolecule damage induced by ionizing radiation. However, proper interpretation of changes in experimental spectra for complex systems is often difficult or impossible, thus Raman spectra calculations based on density functional theory (DFT) provide an invaluable tool as an additional layer of understanding of underlying processes. There are many works that address the problem of basis set dependence for energy and bond length consideration, nevertheless there is still lack of consistent research on basis set influence on Raman spectra intensities for biomolecules. This study fills this gap by investigating of the influence of basis set choice for the interpretation of Raman spectra of the thymine molecule calculated using the DFT/B3LYP framework and comparing these results with experimental spectra. Among 19 selected Pople’s basis sets, the best agreement was achieved using 6-31$+$(d,p), 6-31$++$(d,p) and 6-11$++$G(d,p) sets. Adding diffuse functions or polarized functions for small basis set or use of a medium or large basis set without diffuse or polarized functions is not sufficient to reproduce Raman intensities correctly. The introduction of the diffuse functions ($++$) on hydrogen atoms is not necessary for gas phase calculations. This work serves as a benchmark for further research on the interaction of ionizing radiation with DNA molecules by means of ab initio calculations and Raman spectroscopy. Moreover, this work provides a set of new scaling factors for Raman spectra calculation in the framework of DFT/B3LYP method.

Identification of gene interactions is one of the very well-known and important problems in the field of genetics. However, discovering synergistic gene interactions is a relatively new problem which has been proven to be as significant as the former in genetics. Several approaches have been proposed in this regard and most of them depend upon information theoretic measures. These approaches quantize the gene expression levels, explicitly or implicitly and therefore, may lose information. Here, we have proposed a novel approach for identifying synergistic gene interactions directly from the continuous expression levels, using a minimum spanning tree (MST)-based algorithm. We have used this approach to find pairs of synergistically interacting genes in prostate cancer. The advantages of our method are that it does not need any discretization and it can be extended straightway to find synergistically interacting sets of genes having three or more elements as per the requirement of the situation. We have demonstrated the relevance of the synergistic genes in cancer biology using KEGG pathway analysis and otherwise.

In order to promote the acceptance of cell-based toxicity testings, the accuracy of cytotoxicity test must be determined when compared to in vivo results. Traditional methods of cytotoxicity analysis, such as LC${}_{50}$ (concentration where 50% of the cells are killed) can be problematic since they have been found to vary with time. Technological advances in cytotoxicity testing make it easy to record the dynamic data on changes in cell proliferation, morphology, and damage. To effectively and reasonably analyze the dynamic data, we present a new in vitro toxicity assessed method using the discrete-time Fourier transform (DTFT) which maps the measured cell index from the time domain to the frequency domain. The direct current (DC) component of the DTFT is extracted as a feature which reflects the intensity of cytotoxicity. The smaller the value, the higher the cytotoxicity. Then, a novel toxicity index, as expressed in terms of DC${}_{50}$, is calculated. Results generated with selected test chemicals are compared favorably with data obtained from The Interagency Coordinating Committee on the Validation of Alternative Method (ICCVAM) report concerning the prediction of acute systemic toxicity in rodents. The method can be applied with the standard and high throughput to estimate acute rodent oral toxicity which reduces the number of animals required in subsequent pharmacological/toxicological studies.

The blood–brain barrier (BBB), a highly selective barrier between central nervous system (CNS) and the blood stream, restricts and regulates the penetration of compounds from the blood into the brain. Drugs that affect the CNS interact with the BBB prior to their target site, so the prediction research on BBB permeability is a fundamental and significant research direction in neuropharmacology. In this study, we combed through the available data and then with the help of support vector machine (SVM), we established an experiment process for discovering potential CNS compounds and investigating the mechanisms of BBB permeability of them to advance the research in this field four types of prediction models, referring to CNS activity, BBB permeability, passive diffusion and efflux transport, were obtained in the experiment process. The first two models were used to discover compounds which may have CNS activity and also cross the BBB at the same time; the latter two were used to elucidate the mechanism of BBB permeability of those compounds. Three optimization parameter methods, Grid Search, Genetic Algorithm (GA), and Particle Swarm Optimization (PSO), were used to optimize the SVM models. Then, four optimal models were selected with excellent evaluation indexes (the accuracy, sensitivity and specificity of each model were all above 85%). Furthermore, discrimination models were utilized to study the BBB properties of the known CNS activity compounds in Chinese herbs and this may guide the CNS drug development. With the relatively systematic and quick approach, the application rationality of traditional Chinese medicines for treating nervous system disease in the clinical practice will be improved.

MicroRNAs (miRNAs) are a set of short (21–24 nt) non-coding RNAs that play significant regulatory roles in the cells. Triplet-SVM-classifier and MiPred (random forest, RF) can identify the real pre-miRNAs from other hairpin sequences with similar stem-loop (pseudo pre-miRNAs). However, the 32-dimensional local contiguous structure-sequence can induce a great information redundancy. Therefore, it is essential to develop a method to reduce the dimension of feature space. In this paper, we propose optimal features of local contiguous structure-sequences (OP-Triplet). These features can avoid the information redundancy effectively and decrease the dimension of the feature vector from 32 to 8. Meanwhile, a hybrid feature can be formed by combining minimum free energy (MFE) and structural diversity. We also introduce a neural network algorithm called extreme learning machine (ELM). The results show that the specificity ($S_p$)and sensitivity ($S_n$) of our method are 92.4% and 91.0%, respectively. Compared with Triplet-SVM-classifier, the total accuracy (ACC) of our ELM method increases by 5%. Compared with MiPred (RF) and miRANN, the total accuracy (ACC) of our ELM method increases nearly by 2%. What is more, our method commendably reduces the dimension of the feature space and the training time.