Computational prediction of therapeutic peptides based on graph index.

As therapeutic peptides have been taken into consideration in disease therapy in recent years, many biologists spent time and labor to verify various functional peptides from a large number of peptide sequences. In order to reduce the workload and increase the efficiency of identification of functional proteins, we propose a sequence-based model, q-FP (functional peptide prediction based on the q-Wiener Index), capable of recognizing potentially functional proteins. We extract three types of features by mixing graphic representation and statistical indices based on the q-Wiener index and physicochemical properties of amino acids. Our support-vector-machine-based model achieves an accuracy of 96.71%, 93.34%, 98.40%, and 91.40% for anticancer, virulent, and allergenic proteins datasets, respectively, by using 5-fold cross validation.

Protein Sequence Comparison Based on Physicochemical Properties and the Position-Feature Energy Matrix.

We develop a novel position-feature-based model for protein sequences by employing physicochemical properties of 20 amino acids and the measure of graph energy. The method puts the emphasis on sequence order information and describes local dynamic distributions of sequences, from which one can get a characteristic B-vector. Afterwards, we apply the relative entropy to the sequences representing B-vectors to measure their similarity/dissimilarity. The numerical results obtained in this study show that the proposed methods leads to meaningful results compared with competitors such as Clustal W.

A kernel-based clustering method for gene selection with gene expression data.

Gene selection is important for cancer classification based on gene expression data, because of high dimensionality and small sample size. In this paper, we present a new gene selection method based on clustering, in which dissimilarity measures are obtained through kernel functions. It searches for best weights of genes iteratively at the same time to optimize the clustering objective function. Adaptive distance is used in the process, which is suitable to learn the weights of genes during the clustering process, improving the performance of the algorithm. The proposed algorithm is simple and does not require any modification or parameter optimization for each dataset. We tested it on eight publicly available datasets, using two classifiers (support vector machine, k-nearest neighbor), compared with other six competitive feature selectors. The results show that the proposed algorithm is capable of achieving better accuracies and may be an efficient tool for finding possible biomarkers from gene expression data.

Comparative study of aromaticity in tetraoxa[8]circulenes.

A detailed analysis of the local aromaticity in tetraoxa[8]circulene and its benzo-annelated derivates was performed by means of the energy effect (ef), multicenter delocalization indices (MCI), harmonic oscillator model of aromaticity (HOMA) index, nucleus independent chemical shifts (NICS), and ring current maps. According to the indices employed, the central ring of tetraoxa[8]circulene is antiaromatic, whereas the benzenoid and furanoid rings are aromatic. The calculated current density distribution in tetraoxa[8]circulenes showed that the "annulene within an annulene" model of aromaticity is not applicable for these compounds. It was found that the presence of benzo-annelated rings enhances (respectively diminishes) aromaticity (respectively antiaromaticity) of the five-membered (respectively eight-membered) rings in tetraoxa[8]circulenes. These effects are found to be proportional to the number of benzo-annelated rings. The observed regularities can be rationalized by means of the pairwise energy effect (pef) of cyclic conjugation.

Predicting nucleosome positions in yeast: using the absolute frequency.

Nucleosome is the basic structure of chromatin in eukaryotic cells, and they form the chromatin fiber interconnected by sections of linker DNA. Nucleosome positioning is of great significance for gene transcription regulation. In this paper, we consider the difference of absolute frequency of nucleotides between the nucleosome forming and nucleosome inhibiting sequences. Based on the 2-mer absolute frequency of nucleotides in genome, a new model is constructed to distinguish nucleosome DNA and linker DNA. When used to predict DNA potential for forming nucleosomes in S. cerevisiae, the model achieved a high accuracy of 96.05%. Thus, the model is very useful for predicting nucleosome positioning.

Predicting the classification of transcription factors by incorporating their binding site properties into a novel mode of Chou's pseudo amino acid composition.

Transcription factors (TF) are proteins that control the first step of gene expression, the transcription of DNA into RNA sequences. The mechanism of transcriptional regulatory can be much better understood if the category of transcription factors is known. We developed a new method for predicting the classification of transcription factors by incorporating their binding site properties into a novel mode of Chou's pseudo amino acid composition. The properties include the length of TFBSs for a TF, a new_PWM value, the proportion of not conservative TFBSs, the proportion of nonucleosome of TFBSs, the proportion of conserved-nucleosome of TFBSs, and the GC content of TFBSs. We construct a vector with these properties to represent a TF. Then the vectors which stand for TFs were classified with SVMs. The high accuracy obtained shows that these properties are of great significance for a TF.

Triplet fluoranthenes: aromaticity versus unpaired electrons.

Three fluoranthenes and one substituted fluoranthene, 2,2-dimethyl-2H-dibenzo[cd,k]fluoranthene, were investigated using the unrestricted symmetry-broken and complete active space methods. It was shown that four Kekuléan hydrocarbons are diradicals, implying that their ground state is a triplet. In the energetically less favorable singlet state these hydrocarbons exhibit pronounced diradical character. This occurance is explained with the tendency of the investigated molecules to delocalize their π-electrons. This leads to aromatic stabilization which is stronger than destabilization due to unpaired electrons. Our results for 2,2-dimethyl-2H-dibenzo[cd,k]fluoranthene are in excellent accord with experimental findings of McMaster et al. concerning this compound.

Hydrogen-mediated Stone-Wales isomerization of dicyclopenta[de,mn]anthracene.

The mechanism of transformation of two radicals (R1p and R1i) obtained by addition of a hydrogen atom to an external and internal carbon atom of dicyclopenta[de,mn]anthracene (P1) was investigated. Two pathways were revealed. The first mechanism is a one-step process, whereas the second mechanism includes two transition states and a cyclobutyl intermediate. The formation of R1p and R1i and the homolytic cleavage of the radicals obtained during the isomerization processes were also examined. In both pathways the addition of a hydrogen atom to the internal carbon significantly lowers the activation energy for hydrogen-mediated isomerization of P1 to acefluoranthene. This finding could be explained by the specific electronic structures of the transition states and intermediates participating in the isomerization processes.

Correlations between local aromaticity indices of bipartite conjugated hydrocarbons.

Local aromaticity in rings of conjugated hydrocarbons can be measured in a variety of ways. In the present paper, we concentrate on two of these, namely, EC, i.e., the pi-electron content or pi-electron partition, and ef, the energy effect of cycles. For the central ring in five bipartite conjugated hydrocarbons (anthracene, triphenylene, perylene, coronene, and biphenylene), it was found that EC and ef values are modified in a consistent and predictable manner by annelation with benzenoid rings. Equations are presented for computing EC and ef values for the central ring in terms of three integers representing the numbers of annelated benzenoid rings (A, L, and G for angular, linear, and geminal annelation, respectively). The coefficients of A and G are positive (A > G) and the coefficient of L is negative for benzenoids, but for biphenylene, the situation is reversed for coefficients in the correlation for ef values.

Formation and isomerization of dicyclopenta[de,mn]anthracene. Electronic structure study.

The formation of dicyclopenta[de,mn]anthracene (P1) and its isomerization into dicyclopenta[jk,mn]phenanthrene (P3) was investigated using density functional theory. It was shown that P1 is formed from 1,4-diethynilanthracene, but due to its instability, it undergoes further transformation. This transformation involves rearrangements of some hydrogen atoms and ring contraction/ring expansion process, yielding as a final product the isomer P3. The energies of activation for the P1-->P3 intraconversion show that this reaction is competitive to the other, previously investigated isomerization of P1 into dicyclopenta[de,kl]anthracene (P2). In addition, our investigation shows that the formation of P3 from P1 is energetically more favorable than the formation of P3 from P2. Thus, the presence of the isomer P3 in the reaction mixtures could also be caused by the isomerization of the very unstable isomer P1.

Electronic structure study of thermal intraconversions of some dicyclopenta-fused polycyclic aromatic compounds.

The electronic structure of dicyclopenta[de,mn]anthracene (P1), dicyclopenta[ de,kl]anthracene (P2), and dicyclopenta[jk,mn]phenanthrene (P3) and their mutual isomerization processes are investigated using density functional theory. Two mechanisms for the thermal intraconversion of P1 to P2 were found. The first mechanism occurs via ethynylaceanthrylene (I0), and the second involves a 1,2-hydrogen shift. It is supposed that I0 is initially formed during the flash vacuum pyrolysis experiments, eventually rearranging to P2 on high temperatures. The energetics of the latter mechanism also indicate that P1 isomerizes to P2. The mechanism for a transformation of P2 to P3 is based on a ring contraction/ring conversion process and requires extremely high temperatures. Our investigation is in accord with the experimental results: unsuccessful synthesis of P1, stability of P2 at high temperature, and formation of P3 under extreme temperature regime.

On the cycle-dependence of topological resonance energy.

The topological resonance energy (TRE) was conceived in the 1970s. From the very beginning, it was known that TRE is equal to the collective energy-effect of all cycles present in a conjugated molecule. Also in the 1970s a theory of cyclic conjugation was elaborated, by means of which it was possible to compute the energy-effect ef(Z) of each individual cycle Z present in a conjugated molecule. Yet, the connection between TRE and the ef(Z)-values was, until now, not studied. We now show that TRE and the sum of the ef(Z)-values are closely correlated but that a certain correction needs to be made by taking into account the energy-effects of pairs, triplets, quartets, etc. of cycles.

Variable neighborhood search for extremal graphs. 10. Comparison of irregularity indices for chemical trees.

Chemical graphs, as other ones, are regular if all their vertices have the same degree. Otherwise, they are irregular, and it is of interest to measure their irregularity both for descriptive purposes and for QSAR/QSPR studies. Three indices have been proposed in the literature for that purpose: those of Collatz-Sinogowitz, of Albertson, and of Bell's variance of degrees. We study their properties for the case of chemical trees. Structural conjectures are generated with the system AutoGraphiX, and most of them proved later by mathematical means. Analytical expressions for extremal values are obtained, and extremal graphs are characterized for the two last indices.

Chirality codes and molecular structure.

Some time ago a structure-descriptor, named "chirality code", was put forward [J. Chem. Inf. Comput. Sci. 2001, 41, 369-375], aimed at distinguishing between enantiomers. The chirality code is a sequence of (typically 100) numbers, being equal to the value of a certain "chirality function" at equidistant points within a chosen interval. For molecules of moderate size the chirality function has thousands of peaks (maxima and minima), one for each quartet of atoms. Therefore it looks as if the chirality code cannot provide a faithful representation of the chirality function and thus a faithful representation of the molecular structure. We now show that functional groups present in the molecule result in clusters of near-lying and partially overlapping peaks, whose position in the chirality code is characteristic for the particular functional group. This enables a sound structural interpretation of the chirality code.

Algebraic Kekulé structures of benzenoid hydrocarbons.

An algebraic Kekulé structure of a benzenoid hydrocarbon is obtained from an ordinary Kekulé structure by inscribing into each hexagon the number of pi-electrons which (according to this Kekulé structure) belong to this hexagon. We show that in the case of catafusenes, there is a one-to-one correspondence between ordinary and algebraic Kekulé structures. On the other hand, in the case of perifusenes, one algebraic Kekulé structure may correspond to several ordinary Kekulé structures.