Current Bioinformatics - Volume 9, Issue 4, 2014

In this article, we describe a computational methodology that analyzes the results generated in ligand docking and evaluates the correlation between simulation results and intrinsic characteristics present in the crystallographic structures used in the simulation, such as thermal parameter, resolution, and overall quality of the X-ray diffraction data. We focus our analysis on molecular docking data obtained from application of differential evolution implemented in the program Molegro Virtual Docker. As a protein target, we selected the enzyme 3-dehydroquinate dehydratase (DHQD). This enzyme is part of the shikimate pathway, and a protein target for development of anti-tubercular drugs. We used a set with 20 DHQD crystallographic structures with ligands bound to the active site. In order to identify the best approach to molecular docking, we analyzed crystallographic parameters and looked for correlation between the docking results and structural features present in the protein target. Analysis of docking results helps to identify the best approach to use in ligand docking and identify structural features important for the success of this methodology. Analysis of results generated by ligand docking focused on DHQD made possible to assess the best docking protocol for this enzyme and use this optimized approach in the more computational demanding methodology of virtual screening (VS). We used a data set of natural products to identify structural features important for ligand-binding affinity.

A major objective for constructing gene regulatory networks is to use them as models for designing optimal therapeutic intervention policies within the context of synchronous or asynchronous probabilistic Boolean networks (PBNs). However, most of the previous works focused on the former and only few studied the latter. This paper deals with an optimal control problem in a generalized asynchronous PBN by applying the theory of controlled semi-Markov processes. Specifically, we first describe a control model for a generalized asynchronous PBN as a controlled semi- Markov process model and then solve the corresponding optimal control problem such that the probability that the network reaches a prescribed reward level during a first passage time to some target set is maximal. Numerical examples are then given to demonstrate the effectiveness of the proposed methods.

The prediction of eukaryotic exons is an important topic in bioinformatics. In this paper, a model-independent method based on the singularity detection (SD) algorithm and the three-base periodicity has been developed for predicting exons in DNA sequences of eukaryotes. Using the HMR195 data set, BG570 data set and 200 test data as test sets, we show that, (1) In comparison with the exon prediction by nucleotide distribution (EPND), modified Gabor-wavelet transform (MGWT) and fast Fourier transform plus empirical mode decomposition (FFTEMD) method, the proposed SD method notably improves prediction accuracy of exons, especially short exons or the ability to discern two contiguous short exons disunited by a short intron; (2) The SD method also significantly enhances the performance of the noise suppression in exon prediction over all assessed model-independent methods. The performance of the SD method is evaluated in terms of the signal-to-noise, the approximate correlation, the area under the receiver operating characteristic curve and the accuracy against those of the EPND, MGWT and FFTEMD method over HMR195 data set, BG570 data set and 200 test data. Experimental results demonstrate that the SD method outperforms all assessed model-independent methods with respect to those performance parameters.

As the gap grows tremendously between the numbers of protein sequence and structure, in-silico protein structure prediction plays more and more critical roles in life science. Biennial experiments of Critical Assessment of protein Structure Prediction (CASP), the most authoritative in the field of protein structure prediction, shows that most prediction methods of today are successful in certain aspects, such as comparative modeling. However, incomplete models unexpectedly appear and require further refinement works. Therefore, the present study designed an automated multi-template combination algorithm to perform such refinement works. A total of 59 proteins released during CASP9 prediction season (human group) were selected as experimental targets. Four prediction methods HHpred, Pcons, Modeller, and SAM were used to generate protein models, among which 318 models were incomplete. Automated multitemplate combination algorithm was used in this study to work on each incomplete model, find the missing structures from other models, combine them with the original model, and finally obtain a recombined new model. Our results indicated that the quality of 95.56% of these 318 models was improved after the combination, and the improvement was statistically significant. Therefore, this study provided an effective method to improve the protein model quality.

Basic researchers and clinicians are increasingly aware of the remarkable importance of the fractal approaches in the morphological study of cells and tissues, providing information that can help to understand pathological changes. In our experience, fractal analysis has been able to produce important data on the differential diagnosis in the patient. Here we report new data on the fractal analysis of the bone marrow in myelodysplastic syndromes, a group of hematologic neoplasms characterized by morphological dysplasia, aberrant hematopoiesis, peripheral blood refractory cytopenia, with an increased risk of transformation to acute myeloid leukemia. Ninety cases of Myelodysplastic Syndromes, 20 samples of normal bone marrow, 16 cases of benign hyperplastic bone marrow and 9 cases of acute myeloid leukemia (AML) were studied. In myelodysplastic syndromes, fractal dimension is statistically increased compared with the normal condition, and, moreover, it increases with the severity of the lesion. Statistically, four classes arise. Healthy bone marrow, D = 1.72 ± 0.08, “hyperplasia” and “refractory anemia”, D = 1.79 ± 0.08,” refractory anemia with excess blasts- 1” and “refractory anemia with excess blasts -2”, D = 1.86 ± 0.08 and a fourth group, which represents the most severe condition (HIV-related myelodysplastic syndrome, chronic myelomonocytic leukemia and acute myeloid leukemia), with D = 1.95 ± 0.05, i.e. the complete loss of the diffusion limited aggregation structure that characterizes the normal bone marrow. Fractal analysis appears to be able to add objective information relating to the differential diagnosis in myelodysplastic syndromes.

Similarity analysis of sequences belonging to some protein families indicates the existence of highly variable positions. In this work, a method of interpretation of variability at these positions is presented. The proposed method extends out of the Dot-Matrix method with the possibility of making new analysis of similarity and consideration of physicochemical aspects of variability. These analyses have been made at two consideration levels, i.e. the amino-acid level and codon level. To make this possible, semihomologization and desemihomologization mechanisms have been introduced. As a result, the method of selection of sequences, which best represents the given protein family, has been proposed. A higher frequency of six-codon amino-acids at the highly variable positions has been identified, which can indicate that one-point mutation is the main mechanism of amino-acid codon evolution.

Many complex systems such as biological and social systems can be modeled using graph structures called biological networks and social networks. Instead of studying separately each of the elements composing such complex systems, it is easier to study the networks representing the interactions between the elements of these systems. A commonly known fact in biological and social networks’ analysis is that in most networks some important or influential elements (e.g. essential proteins in PPI networks) are placed in some particular positions in a network. These positions (i.e. vertices) have some particular structural properties. Centrality measures quantify such facts from different points of view. Based on centrality measures the graph elements such as vertices and edges can be ranked from different points of view. Top ranked elements in the graph are supposed to play an important role in the network. This paper presents a comprehensive review of existing different centrality measures and their applications in some biological networks such as Protein-Protein interaction network, residue interaction and gene–gene interaction networks.

Knowledge about the topology of G protein-coupled receptors (GPCRs) can be very useful in predicting diverse range of properties about these proteins, such as function, three dimensional structure, and ligand binding site. Considering that only few GPCRs have known structures, many computational efforts have been carried out to develop methods for predicting their topology. A novel method to predict the location and the length of transmembrane helices in GPCRs was proposed. This method consists of a “one by one” amino acid feature extraction window which makes it possible for the method to learn the amino acid distribution in helical segments of GPCR proteins. It is based on hidden Markov model (HMM) with a specific architecture that takes advantage of Viterbi decoding algorithm and the observed frequency values for adjusting the transition probabilities. The prediction capability of the method was evaluated for per-protein, per-segment and per-residue accuracies on two datasets consisting of 649 (at least one GPCR from each family) and 2898 (all GPCRs) sequences extracted from UniProt database and compared with other commonly used existing methods. It was found that in all three assessments, the prediction accuracies for the new method on the larger dataset, i.e., 2898 GPCRs, were higher than that obtained by other methods. The results showed that our method was able to predict the topology of GPCR proteins without any sequence length limitation with the accuracies of 88.9 % and 87.4% for the small (i.e., 649 GPCRs) and large (i.e., 2898 GPCRs) datasets, respectively. (Availability status: The source code is available upon request from the authors)