<p>Shiv Nadar University is a comprehensive, multidisciplinary, research-focused, and student-centric University offering a full range of academic programs at the undergraduate, postgraduate and doctoral level. With state-of-the-art infrastructure, the University comprises of academic wings, sophisticated labs, international standard sports facilities, amphitheaters, auditorium, conference rooms and smart classrooms. The University&rsquo;s goal is to become internationally recognized for the quality of its research and creative endeavors and their applicability to improving quality of life, generating new insights and expanding the boundaries of human knowledge creativity. Committed to excellence in teaching, research and service, the University aims to serve the higher education needs of India and the world beyond.</p>

Shiv Nadar University

Center for Informatics

Protein function prediction is an important problem in functional genomics. Typically, protein sequences are represented by feature vectors. A major problem of protein datasets that increase the complexity of classification models is their large number of features. The process of drug discovery often involves the use of quantitative structure-activity relationship (QSAR) models to identify chemical structures that could have good inhibitory effects on specific targets and have low toxicity (non-specific activity). QSAR models are regression or classification models used in the chemical and biological sciences. Because of high dimensionality problems, a feature selection problem is imminent. In this study, we thus employ a hybrid Estimation of Distribution Algorithm (EDA) based filter-wrapper methodology to simultaneously extract informative feature subsets and build robust QSAR models. The performance of the algorithm was tested on the benchmark classification challenge datasets obtained from the CoePRa competition platform, developed in 2006. Our results clearly demonstrate the efficacy of a hybrid EDA filter-wrapper algorithm in comparison to the results reported earlier. © 2013 IEEE.

World University root node

2013 IEEE Congress on Evolutionary Computation, CEC 2013

Hybrid feature selection and peptide binding affinity prediction using an EDA based algorithm

The structural, electronic, and magnetic properties of MoS2+x parallelogram shaped platelets having m and n Mo atoms on the adjoining edges have been studied using first principles calculations and by varying m and n from 1 to 6. These platelets have 100\% S coverage on two adjoining edges while 50\% S coverage on the other two edges. The structural stability of the platelets increases with size but the highest occupied molecular orbital-lowest unoccupied molecular orbital (HOMO-LUMO) energy gap in general decreases. There is a triangular metallic corner at the intersection of 100\% S covered edges of the platelets when m = n. On the other hand magnetism is observed on the 50\% S covered edges of the platelets for the sizes greater than that of the (3,4) platelet. The magnetic moments mainly arise from the undercoordinated S(2c) atoms at the 50\% S covered edges rather than from Mo atoms. The criteria for the existence of the magnetic moments on S(2c) atoms are suggested and the electronic structure of the platelets on the edges as well as inside is discussed. © the Owner Societies 2013.

Fulltext

Physical Chemistry Chemical Physics

Atomic structure and edge magnetism in MoS2+x parallelogram shaped platelets

From first principles calculations, we show that (InN)32 nanoparticles favor rock salt structure compared with wurtzite structure in bulk. A phase transition from wurtzite to rock salt structure is known to occur in bulk InN at 12.1 GPa and higher values of pressure for AlN and GaN. However, at the nanoscale we show that this structural transition takes place in (InN)32 without applying pressure. The charge asymmetry value g and cationanion size ratio in InN describe very well this behavior. Similar studies on nanoparticles of AlN and GaN as well as a few other binary compounds such as MgS, AgI, ZnO, and CdSe, however, do not show such a transition. Our results suggest (InN)32 to be a unique candidate as further calculations on a few larger size (InN)n nanoparticles show that a filled cage (two shells) (InN)12at(InN)48 structure of (InN)60 has higher binding energy compared with a rock salt structure of (InN) 64 leading to the conclusion that other 3D structures are likely to become favorable over rock salt structure for larger sizes. © 2013 American Institute of Physics.

Journal of Chemical Physics

Prediction of rock salt structure of (InN)32 nanoparticles from first principles calculations

Molecular phylogeny analysis (MPA) investigates differences in molecular sequences to analyze the evolutionary relationships of organisms or their bio-macromolecules. Currently there are two categories of methods for MPA viz. distance based and character based. All the methods require multiple sequence alignment (MSA) as a prerequisite to MPA and are usually followed by bootstrap analysis. MSA is efficient in terms of time, computations and memory requirement only when size and number of sequences are small. MSA of whole proteomes is computationally intensive and time consuming. In this paper, an attempt is made to propose a new alignment free approach of MPA using Recurrence Quantification Analysis (RQA). The protein sequence is converted into a numeric sequence by assigning a unique score in the form of real numbers to each amino acid and various RQA features are extracted from the numeric sequence. These features are used as co-ordinates for calculating the distance between the sequences using an appropriate distance function. The distance matrix, thus obtained is used for clustering using Neighbor-Joining method. Requirement of time for clustering and tree construction is observed to be significantly reduced in comparison with the alignment-based algorithms. As an example, a test case involving an application of this method for clustering of 59 polyprotein sequences of family Flaviviridae is demonstrated and phylogenetic tree thus obtained was found to be fairly accurate with 95% accuracy. Only 3 misclassifications out of 59 were observed. Thus, the proposed method has potential to be a reasonable alternative for existing MPA methods. © 2013 American Scientific Publishers All rights reserved.

Advanced Science Letters

Phylogenetic clustering of protein sequences using recurrence quantification analysis

Accurate detection of peptides binding to specific Major Histocompatibility Complex Class I (MHC-I) molecules is extremely important for understanding the underlying process of the immune system, as well as for effective vaccine design and developing immunotherapies. Development of learning algorithms and their application for binding predictions have thus speeded up the state-of-the-art in immunological research, in a cost-effective manner. In this work, we propose the application of a hybrid filter-wrapper algorithm employing concepts from the recently developed biogeography based optimization algorithm, in conjunction with SVM and Random Forests for identification of MHC-I binding peptides. In the process, we demonstrate the effectiveness of this evolutionary technique, coupled with weighted heuristics, for the construction of improved prediction models. The experiments have been carried out for the CoEPrA competition datasets (accessible online at: http://www.coepra.org) and the results show a marked improvement over the winner results in some situations and comparably good with regard to others. We thus hope to initiate further research on the application of this new bio-inspired methodology for immunological research. © 2012 Elsevier B.V.

Journal of Immunological Methods

Hybrid biogeography based simultaneous feature selection and MHC class I peptide binding prediction using support vector machines and random forests

Constructing classifier models for gene expression datasets using informative features enhances prediction performance of concerned models. Here, we propose a hybrid Group Search based feature selection (GSO-FS) algorithm which can select relevant gene subsets that can optimally predict cancerous tissue samples. Our experimental results show that the GSO-FS algorithm in combination with SVM classifier performs quite well.

GECCO 2013 - Proceedings of the 2013 Genetic and Evolutionary Computation Conference Companion

Simultaneous gene selection and cancer classification using a hybrid group search optimizer

Pseudo-periodic signals are rampant in biomedical applications but are difficult to analyze. One approach is to compute time domain parameters of each individual cycle in the pseudo-periodic signal. The classic approach requires repeated computation in each cycle, which tends to be either error prone, computationally burdensome, or requires manual effort. We provide a novel combination of the pitch synchronous wavelet transform which when combined with dynamic time warping results in effective quantification of cycles in the pseudo-periodic signal. We demonstrate our application of this method in studying the arterial pulse. The results show that our approach is feasible and effective, and confirms further scope in other applications. © 2013 IEEE.

IEEE EMBS Special Topic Conference on Point-of-Care (POC) Healthcare Technologies: Synergy Towards Better Global Healthcare, PHT 2013

A novel PSWT-DTW approach for analyzing pseudo-periodic signals

Microarray cancer gene expression datasets are high dimensional and thus complex for efficient computational analysis. In this study, we address the problem of simultaneous gene selection and robust classification of cancerous samples by presenting two hybrid algorithms, namely Discrete firefly based Support Vector Machines (DFA-SVM) and DFA-Random Forests (DFA-RF) with weighted gene ranking as heuristics. The performances of the algorithms are then tested using two cancer gene expression datasets retrieved from the Kent Ridge Biomedical Dataset Repository. Our results show that both DFA-SVM and DFA-RF can help in extracting more informative genes aiding to building high performance prediction models. © 2013 Springer.

Advances in Intelligent Systems and Computing

Hybrid firefly based simultaneous gene selection and cancer classification using support vector machines and random forests

We report results of ab initio calculations on Y-doped anion Sin clusters with n = 4-20. Our results suggest two growth behaviors in the intermediate range of n = 8 and 20: (1) There is the formation of linked clusters in which a metal atom links two subclusters and (2) where silicon atoms form a cage structure and the metal atom is inside the cage to produce endohedral cages of silicon clusters. The cluster structures have been identified by comparing the calculated spectra of the electronic states with the photoemission spectra on anion clusters. Our results suggest that in some cases a higher energy isomer may be present in experiments. We report the calculations of the infrared and Raman spectra as well as the dipole moments, electron affinity, and polarizability that could provide other ways of identifying the growth behavior in these clusters. © 2013 American Physical Society.

Physical Review B - Condensed Matter and Materials Physics

Growth behavior, electronic structure, and vibrational properties of SinY anion clusters (n=4-20): Metal atom as linker and endohedral dopant

Using ab initio calculations, we study Nbn and Tan clusters with n=12, 15, and 17 and find superatoms made of transition metals. Nb12 and Ta12 have an empty cage icosahedral structure. Neutral Ta12 has icosahedral symmetry and 2μB magnetic moments while the cage for Nb12 is slightly distorted and has no magnetic moment. These clusters behave like a divalent superatom. Accordingly an oxygen atom interacts exohedrally on a 3-fold site of the icosahedral cage like an MgO molecule and leaves the cage intact while Fe, Ru, and Os atoms can be endohedrally doped in Ta12 to produce electronically closed shell clusters with large highest occupied-lowest unoccupied molecular orbital (HOMO-LUMO) gap. Also isoelectronic Co@X12 and Rh@X12 (X=Nb and Ta) cations have a large HOMO-LUMO gap of up to about 1eV. In these cases the magnetic moment is quenched. On the other hand doping with Ni leaves the magnetic behavior of Ta12 the same. Further anions of Nb15 and Ta15 have a closed electronic shell structure and a singlet ground state with a large HOMO-LUMO gap. We show that isoelectronic Nb14Mo, Nb14W, Ta14Mo, Ta14W, Nb14Mn+, Nb14Re+, Ta14Mn+, and Ta14Re+ all have a large HOMO-LUMO gap of up to about 1.0eV and are magic clusters. Most interestingly Ta17 has a tetrahedrally symmetric Z16 Frank-Kasper polyhedral structure with a large magnetic moment of 5μB. On the other hand the magnetic moment on Nb17 cluster is 1μB. The magnetic moment on Ta17 is surprisingly large and it is counter intuitive as normally the magnetic moments decrease as one goes down in a column in the periodic table but some tantalum clusters behave differently. Doping of the n=17 clusters with Zr at the center leaves Zr@Ta16 magnetic with 2μB and 6μB magnetic moments nearly degenerate while Zr@Nb16 has zero magnetic moment. © 2013 Elsevier B.V.