Search

At present work discusses the current level of computer modeling the relationship structure of organic compounds and drugs and their ability to penetrate the BBB. All descriptors that influence to this permeability within classification and regression QSAR models are generalized and analyzed. The crucial role of H-bond in processes both passive, and active transport across BBB is observed. It is concluded that further research should be focused on interpretation the spatial structure of a full-size P-glycoprotein molecule with high resolution and the creation of QSAR models describing the quantitative relationship between structure and active transport of substances across BBB.

"Peptic ulcers" is the most frequent side effect of non-steroidal anti-inflammatory drugs (NSAIDs). Experimental data indicate that pathogenesis of peptic ulcers cannot be explained only by the inhibition of cyclooxygenases. The knowledge about other molecular mechanisms of action of drugs related with development of peptic ulcers could be useful for design of new safe NSAIDs. However, considerable time and material resources are needed for corresponding experimental research. For simplification of experimental search, we have developed an approach for in silico identification of probable molecular mechanisms of action of drugs related with its side effects. We have created the set of NSAIDs containing 85 substances with data about structures and side effects. The computer program PASS (Prediction of Activity Spectra for Substances) predicting more than 3000 molecular mechanisms of action based on structural formula of substances was used to estimate unknown molecular mechanisms of action for these set of NSAIDs. Statistically significant relationships between predicted molecular mechanisms of action and development of peptic ulcers have been established. We have discovered twenty-six molecular mechanisms of action (two known previously and twenty-four new) which probably related with development of peptic ulcers. By analyzing of Gene Ontology data, signal and metabolic pathways, publications in Medline, we formulated hypotheses about the role of ten molecular mechanisms of action in pathogenesis of peptic ulcer.

Finding and optimization of new HIV-1 integrase inhibitors is an important task due to the drawbacks of the existing antiretroviral therapy caused by the significant adverse effects, high cost of the treatment, and emerging resistance. In this study we used computer-assisted methods for finding and optimization of new HIV-1 integrase inhibitors blocking the 3'-processing reaction. Possibility of application of the specialized version of computer program PASS for searching original inhibitors of 3'-processing reaction among the commercially available samples of chemical compounds, belonging to a new chemical class, is shown. We compared the general pharmacological profiles calculated by PASS for the sets of original integrase inhibitors discovered in the study with those for known inhibitors published in literature. It was shown that the predicted average selectivity is close for the compounds from both sets, while the predicted adverse/toxic actions are less for the original compounds. Differences of pharmacological profiles of original compounds and known from literature inhibitors of 3'-processing pointed out to the essential novelty of our substances. Significantly less deviations from the recommended values for QikProp parameters, which characterized the ADME properties, have been observed for the original compounds than for the integrase inhibitors published in literature. Thus, one may come to the conclusion that the original compounds look more prospective comparing to the earlier known inhibitors of 3'-processing.

We propose an approach to rational design of new polyketides with the required spectrum of biological activity. We developed BioGenPharm software for generation of polyketide combinatorial libraries, prediction of activity spectra for the generated structures and selection of molecules with the required properties on the basis of user defined input parameters and selection criteria. For prediction of polyketide activity spectra we used PASS algorithm (http://www.ibmc.msk.ru/PASS). Validation of PASS prediction ability for polyketides was performed vs. the evaluation set containing 242 natural macrolides from the Dictionary of Natural Products. The mean prediction accuracy was 75,5%. The problem of choice of cutting points for probability of the presence of activity (Pa), which provide optimal combination of such parameters as sensitivity, specificity, concordance was considered. Applicability of the described method has been illustrated by generation of a virtual library of the erythromycin analogues and selection substances for which the probability of hepatotoxic action is low.

A new approach for quantitative structure-activity relationships based on MNA descriptors, fuzzy gradation method and self-consistent regression has been proposed. This approach has been realized in the computer program GUSAR (General Unrestricted Structure Activity Relationships). Our method has been validated on CDK1 inhibitors. Prediction accuracy is comparable with popular methods of 3D QSAR: CoMFA and CoMSIA. However, in contrast to CoMFA and CoMS1A, GUSAR approach does not require information about 3D structure of enzyme and ligand. Application of GUSAR method for heterogeneous training sets has been shown.

HIV-1 integrase is the third enzyme essential for viral replication. It represents an attractive target for new anti-HIV drugs. Diketo acids represent the most prospective class of integrase inhibitors; one of them (S1360) is currently under clinical trials. We prepared the training set containing 90 diketoacid derivatives and performed re-training of computer program PASS. Average accuracy of prediction in LOO cross-validation for HIV-1 integrase inhibitors was shown to be 99%. We selected eight new potential HIV-1 integrase inhibitors among about the 700,000 substances from the databases of commercially available chemicals; anti-integrase activity was found experimentally in two selected compounds.

The G-protein coupled peptide receptors as well as their ligands, endogenous peptides, are involved in regulation of many important physiological processes in the organism and therefore represent attractive targets for pharmaceutical investigation and drug design. With the completion of the human draft genome sequencing, it has become possible to take a comprehensive picture of all genes encoding both peptide receptors and peptides themselves. In the present study a first attempt has been made to carry out a comprehensive analysis of G-protein coupled peptide receptors and their respective endogenous peptide ligands in the human genome. We searched the genome sequence by means of sequential application of standard bioinformatical methods (such as homology search, hierarchical cluster analysis, building of hmm-profiles etc.) with the goal of identifying all the components of peptide ligand/receptor system in the human genome. As a result of this search it was concluded that the probable number of functional peptide receptors in the human genome is 218, and the probable peptide precursors' number is 126 amino acid sequences. These two groups include, respectively, 12 novel G-protein coupled peptide receptors and 10 novel peptide precursors, discovered in the present study. The probable biological functions of newly discovered candidates were determined based on the sequence similarity to the earlier known proteins. Classification of all peptide GPCRs and their ligands based on the ligand specificity was performed for all probable G-protein coupled peptide receptors. The issue of ligand-receptor specificity in the human genome is also discussed.

A new approach to constructing the profiles of protein families is proposed, which uses only structural similarity of amino acid residues. We derived multiple alignments of protein sequences from 3D superpositions of the protein structures and constructed protein family profiles using structural molecular MNA descriptors. MNA (Multilevel Neighborhoods of Atoms) descriptors were developed earlier and are successfully applied for predicting the biological activity in drug-like compounds. In our approach, each aligned position was described by a set of MNA descriptors calculated for each amino acid residue in the alignment column. In this study, we constructed MNA profiles for trypsin, subtilase, and cytochrome P450 protein families and scanned SWISSPROT with some fragments of these profiles. We also calculated the Independence Accuracy of Prediction for each profile fragment. It was shown that the approach developed could be applied to predict protein function.

With the modern molecular biology techniques, it has been possible to detect, isolate and clone biological macromolecules, which could be used as immunogenes in artificial vaccine constructs. In the post-genomic era, the prospective immunogenic components are searched using bionformatic tools and proteomic technologies. Today it is quite realistic to combine the artificial vaccine constructs from the preselected molecular components. Existing computational methods are able to detect the potential immunogenes in genomic sequences, predict their characteristics and subcellular location. The set of methods is designed to predict the T- and B-epitopes that can be used as components of minimal vaccine constructs. The variety of systems for production and delivery of vaccines are developed and tested. These include transgenic plants, bacterial and viral vectors, DNA molecules etc. Several informational resources provide free access to molecular immunology data and deliver services on prediction of antigenic features. Several artificial vaccines have already been launched, but much more preparations are under preclinical and clinical trials. Computer-aided design of vaccines may significantly decrease time and costs required for their development. Modern bioinformatic technologies are now employed for discovery of more effective and potent vaccine.

A software package was designed and used in a detailed study of the contact regions (interfaces) of a large number of protein-protein complexes using the PDB data. It appeared that for about 75% of the complexes the amino acid composition of the subunit surface in the contact region is not essential. Thus one may suggest that, along with the amino acid residues at the interface, the residues in the interior of the globules substantially contribute to protein-protein recognition. Such interactions between quite remote residues are most probably of electrical nature, and are involved in recognition by contributing to the overall electric field created by the protein molecule; the configuration of this field is perhaps the definitive factor of recognition. The overall field of the protein molecule is additively built of the fields created by each constituent residue, and it can be calculated as a sum of the fields created by the protein multipole (aggregate of 'partial' electric charges assigned to every atom of the protein molecule). Preliminary calculations of the remote electrostatic interaction have been performed for ribonuclease subunits in vacuum. The results are indicative of a real possibility that the electric field created by the protein multipole can strongly influence the mutual orientation of molecules before Brownian collisions.

To analyze the interrelationships between the amino acid sequences of the proteins of hepatitis C virus and the functional characteristics of different variants of this virus, a database of protein functional mapping of hepatitis C virus was developed. The database contains amino acid sequences (both full-size and fragmentary) retrieved from accessible databases and experimental data published in literature. The database also contains the results of comparison and treatment of primary data, including alignments and functional regions. On the basis of these data, variable and conservative regions of envelope proteins of hepatitis C virus were revealed. Antigenic and functional maps of structural and nonstructural proteins of the virus were constructed. The most variable region of the envelope protein E2 (HVR1) was analysed. It is assumed that the conservatism of some amino acid positions of HVR1 is related to the functions of this region.

The wide variety of the biological effects of peptides and their high activity are the main reasons for the search for new basic drug structures among them. The most promising compounds can be selected using the PASS computer system (Prediction of Activity Spectra for Substances). This system was originally developed to predict the activities of low-molecular "drug-like" organic compounds. Its predictive capacity is described here by the example of 134 peptides and peptidomimetics with nine known biological activities. Its average predictive power is shown to be approximately 97%. Such an accuracy demonstrates that computer prediction can be applied both to the evaluation of effects and mechanisms of action of endogenous and synthetic peptides and to the screening of new therapeutic agents among the most promising basic structures.

Hemisuccinates, hemiphthalates, acetylsalicylates, cinnamates, and p-methoxycinnamates of lupeol, betulin, and 3-O-acetylbetulin were synthesized via interaction with corresponding acid anhydrides or acid chlorides. A number of betulin esters in position 3 and 28 were shown to exhibit a pronounced hepatoprotective effect similar to that of betulin and silibor. These experimental data were in a good agreement with the computer prediction of their biological activity. Betulin 3,28-bis-hemiphthalate was more effective than carsil in models of experimental hepatitis caused by carbon tetrachloride, tetracycline, and ethanol.

In search of an active alcohol dehydrogenase inhibitor, the structure of which may serve as the basis for a potential drug design, the active site of alcohol dehydrogenase containing NAD and Zn2+ ions was mapped using the method of molecular mechanics. Molecular docking was performed using a number of ligands containing characteristic functional groups: formate ion, ammonia, ammonium ion, methanol, and methylamine. Sites of preferable binding were revealed for each ligand and arranged in order of decreasing energy of binding to the enzyme. A comparison of the predicted ligand-binding sites and the experimental data on the location of water and inhibitor binding sites in the known structures of corresponding alcohol dehydrogenase complexes indicated a coincidence of the complex formation sites, which confirms the validity of the method and provides the requirements for a highly effective inhibitor (the pharmacophore model).

Four amino acid's complementarity hypotheses have been checked using the data on the structure of 122 protein complexes taken from protein Brookhaven Data Bank. No one of hypotheses was conformed at the analysis of the contact region structures of the protein complexes examined.

Specialised system for computer assisted prediction of antiulcer activity of chemical compounds on the basis of their structural formulae is described. Predicted activity spectrum includes antiulcer activity, antisecretory and gastroprotective effects; H2- and M1-receptors blockade, H+, K(+)-ATPase inhibition as the mechanisms of antisecretory effect. It is shown that the average prediction accuracy is 83% in leave-one-out cross validation and 82% for the independent test set. Prediction made for about 300 new chemical compounds provides the selection of 20 potential antiulcer agents, 9 of them were synthesised and tested, 5 compounds have potent antiulcer activity. The discovered antiulcer agents are classified as New Chemical Entry because antiulcer and close effects have not been earlier for any compound of this chemical series.

Computer analysis of hygienic standards in Russia and the USA and of the toxicometric parameters of chemicals characterized by predominantly irritating effects helped create a prognostic model for the calculation of maximum allowable concentrations of chemicals in the air of working zones.

The method of possible conformations calculations for cyclic structures, formed by two (or more) identical parallel polypeptide chains, closed by cross bridges has been elaborated. The algorithm is necessary for rigorous conformational analysis of cyclic regions in immunoglobulin, fibronectin and myosin.

The influence of the dipole-dipole interactions in different secondary structure types is analysed. It is calculated that the values of peptide unit dipole moment projections on the helical axial vector and on the vector, perpendicular to the axis, differ markedly for the alpha-helical conformation (the dipole moment axial component is the maximum and equal to 3,34 debye) and for the extended conformations--beta- and poly-1-proline II type (axial component is the minimum and equal to 0.025 debye; perpendicular component is the maximum). In the course of study it has been suggested that the minimum size of the beta-structure stable fragment is two and of poly-1-proline II type helix--three residues. Taking into account that the dipole moment perpendicular components can be compensated along one complete helical turn, some general conclusions concerning protein folding can be made. The number of folding variants for the alpha-helical proteins is limited, whereas for the beta-structure type proteins they are numerous.

The amino acid composition and sequence in primary structure of 180 proteins have been studied. It is shown that the distribution of amino acid residues is near to a random one, i.e. it is determined by the amino acid composition. The ratio between statistical and unique character of protein primary structures has been discussed. The amino acid sequence is suggested to be unique in fibrous proteins. In contrast the amino acid sequence in globular proteins is a statistical one. The statistical character of amino acids distribution in globular proteins explains the possibility of sensible text generation under the frame shift mutations, deletions and insertions.

Conformational analysis of immunoglobulin G3 (IgG3) hinge region repeating part was performed. In approximation of backbone regular conformations for the case of C2 symmetry (the symmetry axis is perpendicular to S-S bridges and passes through their midpoint) on the basis of total analysis of ring geometry possibilities and estimation of steric conditions, the existence of two conformations with n = 3 (3(10) helix type and polyproline II type) was inferred. A possible role of these conformations in the immunoglobulin functioning was discussed.

Age relationships of integral parameters of heat- and electroconductivity (HC and EC) of human skin were studied. Measurements were performed in 14 body regions (Zakharyin-Gedd zones and opposite points) on 95 practically healthy women 10-63 years old. Cluster analysis of the experimental data has shown that population is evenly distributed in the space of characters. The correlation analysis permits the statement that HC and EC are practically independent parameters. At the same time the correlation coefficients for HC and EC values obtained in the opposite points are rather high. Therefore the integral parameters were introduced for describing the system: mean values (EC and HC), indices of lateral asymmetry IAEC and IAHC), nonhomogeneity indices (NIEC and NIHC). Age relationships of integral parameters and their root-mean-square deviations were computed using smoothing by the moving-average method with Gaussian weight function, whose width was optimized by the minimum of information losses. For EC and NIEC and for root-mean-square deviations of IAHC and NIHC statistically reliable changes with age are observed. They reflect the physiological state of the organism (EC and NIEC being of the leading character). The indicated integral parameters are suggested to be used for estimating the man's state at hormonal reconstructions of the organism.

It is shown that the distribution of identical amino acid residues in the primary structure of 83 non-homologous proteins containing about 14 000 amino acids is near to a casual one, i.e. it is determined by the amino acid composition. A characteristic feature of such distribution is an increased probability of a common grouping of amino acid residues.