-
[show abstract]
[hide abstract]
ABSTRACT: In current computational biology, assigning a protein domain to a fold class is a complicated and controversial task. It can be more challenging in the much harder task of correct identification of protein domain fold pattern solely through using extracted information from protein sequence. To deal with such a challenging problem, the concepts of hyperfold and interlaced folds are introduced for the first time. Each hyperfold is a set of interlaced folds with a centroid fold. These concepts are used to construct a framework for handling the uncertainty involved with the fold classification problem. In this approach, an unknown query protein is assigned to a hyperfold rather than a single fold. Ten different sequence based features are used to predicting the correct hyperfold. This architecture is featured by the Dempster-Shafer theory of evidence through the bodies of evidence and Dempster's rule of combination to combine the hyperfolds. The classification architecture thus developed was applied for identifying protein folds among the 27 famous SCOP fold patterns from a stringent well-known dataset. Compared with the existing predictors tested by the same benchmark dataset, our approach might achieve the better results.
Mathematical biosciences 07/2012; 240(2):148-60. · 1.30 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Proteins interact with each other for performing essential functions of an organism. They change partners to get involved in various processes at different times or locations. Studying variations of protein interactions within a specific process would help better understand the dynamic features of the protein interactions and their functions. We studied the protein interaction network of Saccharomyces cerevisiae (yeast) during the brewing of Japanese sake. In this process, yeast cells are exposed to several stresses. Analysis of protein interaction networks of yeast during this process helps to understand how protein interactions of yeast change during the sake brewing process. We used gene expression profiles of yeast cells for this purpose. Results of our experiments revealed some characteristics and behaviors of yeast hubs and non-hubs and their dynamical changes during the brewing process. We found that just a small portion of the proteins (12.8 to 21.6%) is responsible for the functional changes of the proteins in the sake brewing process. The changes in the number of edges and hubs of the yeast protein interaction networks increase in the first stages of the process and it then decreases at the final stages.
The Journal of Microbiology 12/2011; 49(6):965-73. · 1.10 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Purpose - The purpose of this paper is to design and implement a landmine detection robot (Venus) equipped with three electromagnetic sensors and controlled by ordered weighted averaging (OWA) sensor fusion approach. Higher numbers of detected mines in a fixed time interval and lower total power consumption are the achieved goals of this research. Design/methodology/approach - OWA sensor fusion is exploited for data combination in this paper. Unlike most other landmine detection robots, Venus has three electromagnetic sensors, the positions of which can be adjusted according to the environmental conditions. Also, a novel approach for OWA weight dedication using Gaussian distribution function is applied and the whole idea is evaluated practically in several randomly mined fields. Finally, for better evaluation, performance of Venus is compared with the other two landmine detection robots. Findings - The simulation and experimental results proved that in a predetermined interval of time, not only total energy consumption is reduced, but also by expanding the surface and the depth of influence of electromagnetic waves, the number of detected mines is considerably raised. Social implications - In contrast to the regular demining process, which is relatively expensive and complicated, the landmine detection method proposed in this research is surprisingly simple, cost effective, and efficient. Therefore, it may be attractive for every company or organization in this field of research. Originality/value - The paper describes research which implements and evaluates a novel control approach based on OWA sensor fusion method, a new way of using Gaussian distribution function for determining OWA weights, and also an adaptive physical configuration for sensors based on environmental conditions.
Sensor Review 09/2011; 31(4):328-340. · 0.60 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: In this letter, we propose a learning system, active decision fusion learning (ADFL), for active fusion of decisions. Each decision maker, referred to as a local decision maker, provides its suggestion in the form of a probability distribution over all possible decisions. The goal of the system is to learn the active sequential selection of the local decision makers in order to consult with and thus learn the final decision based on the consultations. These two learning tasks are formulated as learning a single sequential decision-making problem in the form of a Markov decision process (MDP), and a continuous reinforcement learning method is employed to solve it. The states of this MDP are decisions of the attended local decision makers, and the actions are either attending to a local decision maker or declaring final decisions. The learning system is punished for each consultation and wrong final decision and rewarded for correct final decisions. This results in minimizing the consultation and decision-making costs through learning a sequential consultation policy where the most informative local decision makers are consulted and the least informative, misleading, and redundant ones are left unattended. An important property of this policy is that it acts locally. This means that the system handles any nonuniformity in the local decision maker's expertise over the state space. This property has been exploited in the design of local experts. ADFL is tested on a set of classification tasks, where it outperforms two well-known classification methods, Adaboost and bagging, as well as three benchmark fusion algorithms: OWA, Borda count, and majority voting. In addition, the effect of local experts design strategy on the performance of ADFL is studied, and some guidelines for the design of local experts are provided. Moreover, evaluating ADFL in some special cases proves that it is able to derive the maximum benefit from the informative local decision makers and to minimize attending to redundant ones.
Neural Computation 02/2011; 23(2):558-91. · 1.88 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: Protein function is related to its chemical reaction to the surrounding environment including other proteins. On the other hand, this depends on the spatial shape and tertiary structure of protein and folding of its constituent components in space. The correct identification of protein domain fold solely using extracted information from protein sequence is a complicated and controversial task in the current computational biology. In this article a combined classifier based on the information content of extracted features from the primary structure of protein has been introduced to face this challenging problem. In the first stage of our proposed two-tier architecture, there are several classifiers each of which is trained with a different sequence based feature vector. Apart from the application of the predicted secondary structure, hydrophobicity, van der Waals volume, polarity, polarizability, and different dimensions of pseudo-amino acid composition vectors in similar studies, the position specific scoring matrix (PSSM) has also been used to improve the correct classification rate (CCR) in this study. Using K-fold cross validation on training dataset related to 27 famous folds of SCOP, the 28 dimensional probability output vector from each evidence theoretic K-NN classifier is used to determine the information content or expertness of corresponding feature for discrimination in each fold class. In the second stage, the outputs of classifiers for test dataset are fused using Sugeno fuzzy integral operator to make better decision for target fold class. The expertness factor of each classifier in each fold class has been used to calculate the fuzzy integral operator weights. Results make it possible to provide deeper interpretation about the effectiveness of each feature for discrimination in target classes for query proteins.
Computational biology and chemistry 12/2010; 35(1):1-9. · 1.37 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: It has been understood that biological networks have modular organizations which are the sources of their observed complexity. Analysis of networks and motifs has shown that two types of hubs, party hubs and date hubs, are responsible for this complexity. Party hubs are local coordinators because of their high co-expressions with their partners, whereas date hubs display low co-expressions and are assumed as global connectors. However there is no mutual agreement on these concepts in related literature with different studies reporting their results on different data sets. We investigated whether there is a relation between the biological features of Saccharomyces Cerevisiae's proteins and their roles as non-hubs, intermediately connected, party hubs, and date hubs. We propose a classifier that separates these four classes.
We extracted different biological characteristics including amino acid sequences, domain contents, repeated domains, functional categories, biological processes, cellular compartments, disordered regions, and position specific scoring matrix from various sources. Several classifiers are examined and the best feature-sets based on average correct classification rate and correlation coefficients of the results are selected. We show that fusion of five feature-sets including domains, Position Specific Scoring Matrix-400, cellular compartments level one, and composition pairs with two and one gaps provide the best discrimination with an average correct classification rate of 77%.
We study a variety of known biological feature-sets of the proteins and show that there is a relation between domains, Position Specific Scoring Matrix-400, cellular compartments level one, composition pairs with two and one gaps of Saccharomyces Cerevisiae's proteins, and their roles in the protein interaction network as non-hubs, intermediately connected, party hubs and date hubs. This study also confirms the possibility of predicting non-hubs, party hubs and date hubs based on their biological features with acceptable accuracy. If such a hypothesis is correct for other species as well, similar methods can be applied to predict the roles of proteins in those species.
BMC Systems Biology 01/2010; 4:172. · 3.15 Impact Factor
-
[show abstract]
[hide abstract]
ABSTRACT: The first question answered in this paper is whether or not learning attention control in the decision space is feasible and
how to develop an online as well as interactive learning approach for such control in this space, in case of feasibility.
Here, decision space is formed by the decision vector of the agents each has allowed to dynamically observe just a subset
of all available sensors. Attention control in this new space means active and dynamic selection of these decision agents
to contribute in making final decision. The second debate is verifying the advantages of attention control in decision space
over that in perceptual space. According to the tight coupling of attention control and motor action selection, in order to
answer above mentioned questions, attention control and motor action selection are formulated in a unified optimization problem
and reinforcement learning is utilized to solve it. In addition to the theoretic comparison of learning attention control
in perceptual and decision space in terms of computational complexity, two proposed approaches are tested on a simple traffic
sign recognition task.
02/2009: pages 242-256;
-
21st IEEE Canadian Conference on Electrical and Computer Engineering; 05/2008
-
3rd IFAC Workshop on Advanced Fuzzy and Neural Control, University of Valenciennes et du Hainaut Cambrésis, France; 10/2007
-
The International Colloquium on Information Fusion (ICIF), Xi’an, China; 08/2007
-
5th IFAC Symposium on Advances in Automotive Control, California, USA; 08/2007
-
-
[show abstract]
[hide abstract]
ABSTRACT: The proposed approach is based on the classical model of Mixture-of-Experts and tries to concurrently learn two tightly coupled issues. As the main goal, it learns the optimal classification and at the same time, it learns the best sequence of council with previously designed local decision experts to reach the former optimal classification strat-egy. Local experts are in fact local classifiers who have learned the sub-optimal decision making based on just a portion of the whole feature space. The methodology is that in the first stage we generate different feature spaces by binning the features according to their potential relevance and then randomly selecting from the bins. At the second stage, we train a classifier for each of the resulting feature subsets. Finally, we use a continuous Q-learning variant for learning a combiner for the predictions of these classifiers which is the key contribution of the paper. Actually, the meta-learner in the last stage learns to combine different global models, each induced from a different feature subset. The domain of medical diagnosis (specifically Arrhythmia recognition) by using UCI datasets is opted as the benchmark. The acquired classification rate certifies that the proposed approach is quite comparable with the results have been reported so far. Moreover, this recognition is achieved by as few consultations as possible which is another key different merit for our approach.
-
