Recently, many concerns are paid for dual action drugs such as ACE/NEP dual inhibitors which have two different biological activities. To identify multiple active drugs by supervised learning approach, a multi-label classification technique is required. In the present work, we investigated the classification of antihypertensive drugs including ACE/NEP dual inhibitors using support vector machines (SVMs). Biological activity data of the drugs were taken from the MDDR database and they were employed for the computational trial for the training of the SVM classifiers. Structural feature representation of each drug molecule was based on topological fragment spectra (TFS) method. The obtained classifiers were tested for finding ACE/NEP dual inhibitors. The result suggests that the TFS-based SVM classifiers are useful for finding multiple active drugs such as ACE/NEP dual inhibitors.
{"title":"Identification of the Dual Action Antihypertensive Drugs Using TFS-Based Support Vector Machines","authors":"Kentaro Kawai, Yoshimasa Takahashi","doi":"10.1273/CBIJ.10.E_1","DOIUrl":"https://doi.org/10.1273/CBIJ.10.E_1","url":null,"abstract":"Recently, many concerns are paid for dual action drugs such as ACE/NEP dual inhibitors which have two different biological activities. To identify multiple active drugs by supervised learning approach, a multi-label classification technique is required. In the present work, we investigated the classification of antihypertensive drugs including ACE/NEP dual inhibitors using support vector machines (SVMs). Biological activity data of the drugs were taken from the MDDR database and they were employed for the computational trial for the training of the SVM classifiers. Structural feature representation of each drug molecule was based on topological fragment spectra (TFS) method. The obtained classifiers were tested for finding ACE/NEP dual inhibitors. The result suggests that the TFS-based SVM classifiers are useful for finding multiple active drugs such as ACE/NEP dual inhibitors.","PeriodicalId":40659,"journal":{"name":"Chem-Bio Informatics Journal","volume":"74 1","pages":"41-51"},"PeriodicalIF":0.3,"publicationDate":"2009-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86329714","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We have improved a modified charge equilibration (MQEq) method for calculating the geometry-dependent distribution of atomic charges. In this paper, Ohno-Klopman, Ohno and DasGupta-Huzinaga equations are adopted to express the shielding effect, and the calculated atomic charges with these MQEq methods are in good agreement with those by the HF/6-31G(d,p) calculations for several organic molecules. These MQEq methods would be useful to estimate the charge distribution for large molecules.
{"title":"Molecular geometry-dependent atomic charge calculation with modified charge equilibration method","authors":"K. Aoki, S. Tanaka, T. Nakano","doi":"10.1273/CBIJ.9.30","DOIUrl":"https://doi.org/10.1273/CBIJ.9.30","url":null,"abstract":"We have improved a modified charge equilibration (MQEq) method for calculating the geometry-dependent distribution of atomic charges. In this paper, Ohno-Klopman, Ohno and DasGupta-Huzinaga equations are adopted to express the shielding effect, and the calculated atomic charges with these MQEq methods are in good agreement with those by the HF/6-31G(d,p) calculations for several organic molecules. These MQEq methods would be useful to estimate the charge distribution for large molecules.","PeriodicalId":40659,"journal":{"name":"Chem-Bio Informatics Journal","volume":"17 1","pages":"30-40"},"PeriodicalIF":0.3,"publicationDate":"2009-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74707116","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We conducted a docking efficiency validation of ArgusLab, a free docking software program. In this study, the calculated binding free energies of protein-ligand complexes by scoring functions were compared with experimental binding affinities. Correlations between the calculated and experimental data were evaluated for 11 ArgusLab settings and compared. Our results indicate that ArgusLab is useful for virtual screening and the weight of van der Waals interactions are unimportant for binding free energy calculations using this software.
{"title":"Validation of ArgusLab Efficiencies for Binding Free Energy Calculations","authors":"A. Oda, O. Takahashi","doi":"10.1273/CBIJ.9.52","DOIUrl":"https://doi.org/10.1273/CBIJ.9.52","url":null,"abstract":"We conducted a docking efficiency validation of ArgusLab, a free docking software program. In this study, the calculated binding free energies of protein-ligand complexes by scoring functions were compared with experimental binding affinities. Correlations between the calculated and experimental data were evaluated for 11 ArgusLab settings and compared. Our results indicate that ArgusLab is useful for virtual screening and the weight of van der Waals interactions are unimportant for binding free energy calculations using this software.","PeriodicalId":40659,"journal":{"name":"Chem-Bio Informatics Journal","volume":"197 ","pages":"52-61"},"PeriodicalIF":0.3,"publicationDate":"2009-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1273/CBIJ.9.52","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72542530","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Hideki Tanizawa, M. Taniguchi, Ganga D. Ghimire, S. Mitaku
A prediction system for identifying the region of flexible regions of the coiled coil was developed to determine the bending positions of the myosin rods using atomic force microscopy (AFM) and to analyze the molecular structures of proteins containing coiled coils. The prediction system comprises two modules: identification of heptad break points and prediction of fragile points in the coiled coil due to the hydrophilic core or hydrophobic outfield region. Here, we investigated the myosin rods using this prediction system. The results of AFM imaging showed four main flexible regions in a single myosin rod and of the 17 possible fragile points predicted, 16 were located in the four experimental bending regions. Next, we examined the enhanced fluctuation around these predicted fragile points using the B-factor for the three dimensional structure of coiled coil proteins from the SCOP database and found that the fluctuations in the hydrophilic core regions were significantly larger than those in the regions of the normal coiled coil. In contrast, the fluctuations in the hydrophobic outfield regions were reduced, suggesting a structural change of the coiled coils to balance these regions. Thus, the dynamic changes in the structure of the coiled coils around the fragile points may be related to the biological functions of the proteins. The prediction tool which developed in this work was incorporated in the SOSUIcoil system which predicts the coiled coil regions.
{"title":"Prediction of fragile points of coiled coils","authors":"Hideki Tanizawa, M. Taniguchi, Ganga D. Ghimire, S. Mitaku","doi":"10.1273/CBIJ.9.12","DOIUrl":"https://doi.org/10.1273/CBIJ.9.12","url":null,"abstract":"A prediction system for identifying the region of flexible regions of the coiled coil was developed to determine the bending positions of the myosin rods using atomic force microscopy (AFM) and to analyze the molecular structures of proteins containing coiled coils. The prediction system comprises two modules: identification of heptad break points and prediction of fragile points in the coiled coil due to the hydrophilic core or hydrophobic outfield region. Here, we investigated the myosin rods using this prediction system. The results of AFM imaging showed four main flexible regions in a single myosin rod and of the 17 possible fragile points predicted, 16 were located in the four experimental bending regions. Next, we examined the enhanced fluctuation around these predicted fragile points using the B-factor for the three dimensional structure of coiled coil proteins from the SCOP database and found that the fluctuations in the hydrophilic core regions were significantly larger than those in the regions of the normal coiled coil. In contrast, the fluctuations in the hydrophobic outfield regions were reduced, suggesting a structural change of the coiled coils to balance these regions. Thus, the dynamic changes in the structure of the coiled coils around the fragile points may be related to the biological functions of the proteins. The prediction tool which developed in this work was incorporated in the SOSUIcoil system which predicts the coiled coil regions.","PeriodicalId":40659,"journal":{"name":"Chem-Bio Informatics Journal","volume":"105 1","pages":"12-29"},"PeriodicalIF":0.3,"publicationDate":"2009-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74476388","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Chemical modification of an active drug with promoiety to a prodrug (“prodrugging”) is a way to improve the pharmacokinetic characteristics of an active drug in the body; however, no kinetic principles have been proposed to design orally effective prodrugs to overcome biological membrane barriers. Therefore, based on a previously reported kinetic model of drug absorption [Mizuma et al., J Pharm. Sci. 85, 854 (1996)], conditional equations for the kinetic strategy of prodrugging were derived. Conditional equations contain terms of uptake (influx) and efflux transport of the prodrug and drug, and the metabolism of the prodrug to drug. Thereby, kinetic classification and criteria for effective membrane-permeable prodrugs are shown as a decision tree with conditional equations. The first point in the kinetic classification and criteria is the uptake process; second, the efflux process is vetted; finally, the metabolic process is elucidated. In some cases, metabolism is not a factor in the better absorption of prodrugs than active drugs. Experiments corresponding to particular processes were proposed, and are applicable to the design of prodrugs not only for intestinal absorption, but also for biological membrane permeation.
将具有促进作用的活性药物化学修饰为前药(“前药”)是一种改善活性药物在体内药代动力学特性的方法;然而，目前还没有提出动力学原理来设计口服有效的前药来克服生物膜屏障。因此，基于先前报道的药物吸收动力学模型[Mizuma et al.， J Pharm.]科学学报，85,854(1996)]，导出了前移动力学策略的条件方程。条件方程包含前药和药物的摄取(内流)和外排运输，以及前药到药物的代谢。因此，有效的膜渗透性前药的动力学分类和标准被表示为具有条件方程的决策树。动力学分类和标准的第一点是吸收过程;其次，对外排过程进行审查;最后，对代谢过程进行了阐述。在某些情况下，代谢不是前药比活性药更好吸收的一个因素。提出了与特定工艺相对应的实验，不仅适用于肠道吸收，也适用于生物膜渗透的前药设计。
{"title":"Pharmacokinetic strategy for designing orally effective prodrugs overcoming biological membrane barriers: proposal of kinetic classification and criteria for membrane-permeable prodrug-likeness","authors":"T. Mizuma","doi":"10.1273/CBIJ.8.25","DOIUrl":"https://doi.org/10.1273/CBIJ.8.25","url":null,"abstract":"Chemical modification of an active drug with promoiety to a prodrug (“prodrugging”) is a way to improve the pharmacokinetic characteristics of an active drug in the body; however, no kinetic principles have been proposed to design orally effective prodrugs to overcome biological membrane barriers. Therefore, based on a previously reported kinetic model of drug absorption [Mizuma et al., J Pharm. Sci. 85, 854 (1996)], conditional equations for the kinetic strategy of prodrugging were derived. Conditional equations contain terms of uptake (influx) and efflux transport of the prodrug and drug, and the metabolism of the prodrug to drug. Thereby, kinetic classification and criteria for effective membrane-permeable prodrugs are shown as a decision tree with conditional equations. The first point in the kinetic classification and criteria is the uptake process; second, the efflux process is vetted; finally, the metabolic process is elucidated. In some cases, metabolism is not a factor in the better absorption of prodrugs than active drugs. Experiments corresponding to particular processes were proposed, and are applicable to the design of prodrugs not only for intestinal absorption, but also for biological membrane permeation.","PeriodicalId":40659,"journal":{"name":"Chem-Bio Informatics Journal","volume":"125 1","pages":"25-32"},"PeriodicalIF":0.3,"publicationDate":"2008-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89135154","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Akihiko Kishimoto, Yukio Ono, K. Murakawa, T. Ishibashi, A. Wakamatsu, K. Kanehori, N. Nomura, T. Isogai, M. Yohda, S. Sugano
In the Full-length Human cDNA Sequencing Project, 30,160 cDNA were sequenced. Among them, our group performed sequencing of 3,588 cDNAs, mainly using the primer walking method. The sequences achieved an average Phrap score of 76, which means the average of expected sequence accuracy was 99.9999975%, by sequencing of both strands with the criterion of a Phrap score over 30. In spite of the extremely high sequence reliability, we met with difficulty in sequencing 52 cDNAs, which are termed undecipherable cDNAs. cDNAs of long repeats were considered as a possible source of sequencing difficulty; their maximum repeat length sequenced by the primer walking method was 530 bp, without using the random method, and 81% of long repeat sequences remained in the ORFs. In single repeat regions, the insertion/deletion rates were much larger than in the usual regions. The fraction of SINE/Alu repeats in the cDNAs was 5.4%, half of the fraction of the human genome. The fraction of SINE/Alu in undecipherable cDNAs was up to 10%, the same level of the human genome.
{"title":"Classification and characterization of human full-length cDNA clones that are difficult to sequence","authors":"Akihiko Kishimoto, Yukio Ono, K. Murakawa, T. Ishibashi, A. Wakamatsu, K. Kanehori, N. Nomura, T. Isogai, M. Yohda, S. Sugano","doi":"10.1273/CBIJ.8.1","DOIUrl":"https://doi.org/10.1273/CBIJ.8.1","url":null,"abstract":"In the Full-length Human cDNA Sequencing Project, 30,160 cDNA were sequenced. Among them, our group performed sequencing of 3,588 cDNAs, mainly using the primer walking method. The sequences achieved an average Phrap score of 76, which means the average of expected sequence accuracy was 99.9999975%, by sequencing of both strands with the criterion of a Phrap score over 30. In spite of the extremely high sequence reliability, we met with difficulty in sequencing 52 cDNAs, which are termed undecipherable cDNAs. cDNAs of long repeats were considered as a possible source of sequencing difficulty; their maximum repeat length sequenced by the primer walking method was 530 bp, without using the random method, and 81% of long repeat sequences remained in the ORFs. In single repeat regions, the insertion/deletion rates were much larger than in the usual regions. The fraction of SINE/Alu repeats in the cDNAs was 5.4%, half of the fraction of the human genome. The fraction of SINE/Alu in undecipherable cDNAs was up to 10%, the same level of the human genome.","PeriodicalId":40659,"journal":{"name":"Chem-Bio Informatics Journal","volume":"23 1","pages":"1-13"},"PeriodicalIF":0.3,"publicationDate":"2008-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85575991","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The coiled coil structure in proteins is characterized by heptad repeats of hydrophobic amino acids, but many breaks of the heptad repeats are observed within coiled coil regions which are the main cause of the errors in the currently used prediction systems for coiled coils. For understanding the characteristics of coiled coils including heptad breaks, the features of coiled coils were studies focusing on three problems: (1) the determination of appropriate register for the breaks in heptad repeat regions, (2) the discriminations of coiled coil regions using physical properties of amino acid segments which have heptad repeats and (3) the elucidation of the structural difference among several types of heptad breaks. Appropriate registers of heptad repeats and breaks were determined by two steps: first, the typical template of heptad repeats was applied to amino acid sequences and then several types of template for heptad breaks were applied to the segments around inconsistent points of the heptad repeat regions, leading to the most appropriate registers. Then, the coiled coil regions were discriminated from other types of regions by the canonical discriminant analysis, using ten parameters (three physicochemical properties and seven number densities of amino acids). The novel coiled coil prediction system SOSUIcoil showed better performances of coiled coil prediction than other prediction systems. Furthermore, the structures of segments around the heptad breaks were analyzed, indicating that some types of heptad breaks tend to form coiled coil structure whereas the other types are at the end of coiled coils.
{"title":"A high performance prediction system of coiled coil domains containing heptad breaks: SOSUIcoil","authors":"Hideki Tanizawa, Ganga D. Ghimire, S. Mitaku","doi":"10.1273/CBIJ.8.96","DOIUrl":"https://doi.org/10.1273/CBIJ.8.96","url":null,"abstract":"The coiled coil structure in proteins is characterized by heptad repeats of hydrophobic amino acids, but many breaks of the heptad repeats are observed within coiled coil regions which are the main cause of the errors in the currently used prediction systems for coiled coils. For understanding the characteristics of coiled coils including heptad breaks, the features of coiled coils were studies focusing on three problems: (1) the determination of appropriate register for the breaks in heptad repeat regions, (2) the discriminations of coiled coil regions using physical properties of amino acid segments which have heptad repeats and (3) the elucidation of the structural difference among several types of heptad breaks. Appropriate registers of heptad repeats and breaks were determined by two steps: first, the typical template of heptad repeats was applied to amino acid sequences and then several types of template for heptad breaks were applied to the segments around inconsistent points of the heptad repeat regions, leading to the most appropriate registers. Then, the coiled coil regions were discriminated from other types of regions by the canonical discriminant analysis, using ten parameters (three physicochemical properties and seven number densities of amino acids). The novel coiled coil prediction system SOSUIcoil showed better performances of coiled coil prediction than other prediction systems. Furthermore, the structures of segments around the heptad breaks were analyzed, indicating that some types of heptad breaks tend to form coiled coil structure whereas the other types are at the end of coiled coils.","PeriodicalId":40659,"journal":{"name":"Chem-Bio Informatics Journal","volume":"21 1","pages":"96-111"},"PeriodicalIF":0.3,"publicationDate":"2008-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80866636","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Ganga D. Ghimire, Hideki Tanizawa, M. Sonoyama, S. Mitaku
G-protein coupled receptors (GPCRs) bind with G-proteins upon activation by ligands. Understanding the mechanisms of specific binding between GPCRs and G-proteins is one of the most important issues in bioinformatics research. In this study, the physical properties of various regions were analyzed in order to classify GPCRs by G-protein family and to better understand binding specificity. We focused on cytoplasmic loops (IL1, IL2 and N/C-terminus of IL3), extracellular loops (NTL, EL1 and N/C-terminus of EL2) and cytoplasmic termini of transmembrane helices, except for helices that connect to C-terminus loops. The distribution of hydrophobicity, charge density, lysine and arginine densities, and loop length enabled discrimination of GPCRs with more than 90% accuracy.
{"title":"Physicochemical properties of GPCR amino acid sequences for understanding GPCR-G-protein coupling","authors":"Ganga D. Ghimire, Hideki Tanizawa, M. Sonoyama, S. Mitaku","doi":"10.1273/CBIJ.8.49","DOIUrl":"https://doi.org/10.1273/CBIJ.8.49","url":null,"abstract":"G-protein coupled receptors (GPCRs) bind with G-proteins upon activation by ligands. Understanding the mechanisms of specific binding between GPCRs and G-proteins is one of the most important issues in bioinformatics research. In this study, the physical properties of various regions were analyzed in order to classify GPCRs by G-protein family and to better understand binding specificity. We focused on cytoplasmic loops (IL1, IL2 and N/C-terminus of IL3), extracellular loops (NTL, EL1 and N/C-terminus of EL2) and cytoplasmic termini of transmembrane helices, except for helices that connect to C-terminus loops. The distribution of hydrophobicity, charge density, lysine and arginine densities, and loop length enabled discrimination of GPCRs with more than 90% accuracy.","PeriodicalId":40659,"journal":{"name":"Chem-Bio Informatics Journal","volume":"41 1","pages":"49-57"},"PeriodicalIF":0.3,"publicationDate":"2008-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79206474","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
N. Takano, K. Umezawa, Jinzen Ikebe, Yuki Sonobe, Ryosuke Yagisawa, Ito Junichi, N. Hamasaki, D. Mitomo, H. Miyagawa, A. Yamagishi, J. Higo
We did a molecular dynamics simulation of a system consisting of a peptide and a protein in explicit solvent to study biomolecular approach process. In the initial structure of simulation, the minimum inter-biomolecular distance was 30 A. During the simulation, the biomolecules approached and contacted to each other. In spite of diffusive motions of water molecules, the orientations of water molecules tended to order in the inter-biomolecular zone showing coherent spatial patterns (solvent site-dipole field) of the ordering. The degree of ordering was synchronized well with the inter-biomolecular distance. This result strongly suggests that the biomolecules distant to each other can interact via the solvent site-dipole field. The effective range for the coherent ordering (i.e., the interaction range via the solvent site-dipole field) was larger than 20 A. A bridge-like structure of the solvent orientational ordering connected the two biomolecules. Biological and physicochemical significance of the ordering is discussed.
{"title":"Solvent Site-Dipole Field Accompanying Protein-Ligand Approach Process","authors":"N. Takano, K. Umezawa, Jinzen Ikebe, Yuki Sonobe, Ryosuke Yagisawa, Ito Junichi, N. Hamasaki, D. Mitomo, H. Miyagawa, A. Yamagishi, J. Higo","doi":"10.1273/CBIJ.8.14","DOIUrl":"https://doi.org/10.1273/CBIJ.8.14","url":null,"abstract":"We did a molecular dynamics simulation of a system consisting of a peptide and a protein in explicit solvent to study biomolecular approach process. In the initial structure of simulation, the minimum inter-biomolecular distance was 30 A. During the simulation, the biomolecules approached and contacted to each other. In spite of diffusive motions of water molecules, the orientations of water molecules tended to order in the inter-biomolecular zone showing coherent spatial patterns (solvent site-dipole field) of the ordering. The degree of ordering was synchronized well with the inter-biomolecular distance. This result strongly suggests that the biomolecules distant to each other can interact via the solvent site-dipole field. The effective range for the coherent ordering (i.e., the interaction range via the solvent site-dipole field) was larger than 20 A. A bridge-like structure of the solvent orientational ordering connected the two biomolecules. Biological and physicochemical significance of the ordering is discussed.","PeriodicalId":40659,"journal":{"name":"Chem-Bio Informatics Journal","volume":"40 1","pages":"14-24"},"PeriodicalIF":0.3,"publicationDate":"2008-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82634055","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Michio Koyama, K. Hasegawa, Masamoto Arakawa, K. Funatsu
In the field of drug discovery, high-throughput screening (HTS) is widely used to identify new lead compounds. A considerable number of hit compounds, however, will subsequently be found to have low activities when their inhibitory activities are measured more precisely. Such compounds are called false positives. For a more efficient selection of lead compounds, virtual screening methods with QSAR models have been investigated, but no definitive solutions have been found. In this study, we propose an effective method to identify lead compounds. The proposed method is based on rough set theory (RST), which is a mathematical tool for depicting the uncertainty and vagueness of knowledge. The essential parts of RST are the construction of reducts, which are minimal subsets of variables to distinguish samples, and the extraction of rules using their reducts. By applying RST to the QSAR study of monoamine oxidase (MAO) inhibitors, we extracted several rules for identifying lead compounds. First, 3D-structures of MAO inhibitors were generated uniformly by CORINA, and chemical descriptors were calculated by the Volsurf method. Finally, three unique rules were extracted by using RST. It is found that the each rule is chemically reasonable and compatible with previous studies. Furthermore, the predictive power of RST was also proved by comparison with partial least squares (PLS) and decision tree (DT). These results demonstrate the usefulness of our method.
{"title":"Application of Rough Set Theory to High Throughput Screening Data for Rational Selection of Lead Compounds","authors":"Michio Koyama, K. Hasegawa, Masamoto Arakawa, K. Funatsu","doi":"10.1273/CBIJ.8.85","DOIUrl":"https://doi.org/10.1273/CBIJ.8.85","url":null,"abstract":"In the field of drug discovery, high-throughput screening (HTS) is widely used to identify new lead compounds. A considerable number of hit compounds, however, will subsequently be found to have low activities when their inhibitory activities are measured more precisely. Such compounds are called false positives. For a more efficient selection of lead compounds, virtual screening methods with QSAR models have been investigated, but no definitive solutions have been found. In this study, we propose an effective method to identify lead compounds. The proposed method is based on rough set theory (RST), which is a mathematical tool for depicting the uncertainty and vagueness of knowledge. The essential parts of RST are the construction of reducts, which are minimal subsets of variables to distinguish samples, and the extraction of rules using their reducts. By applying RST to the QSAR study of monoamine oxidase (MAO) inhibitors, we extracted several rules for identifying lead compounds. First, 3D-structures of MAO inhibitors were generated uniformly by CORINA, and chemical descriptors were calculated by the Volsurf method. Finally, three unique rules were extracted by using RST. It is found that the each rule is chemically reasonable and compatible with previous studies. Furthermore, the predictive power of RST was also proved by comparison with partial least squares (PLS) and decision tree (DT). These results demonstrate the usefulness of our method.","PeriodicalId":40659,"journal":{"name":"Chem-Bio Informatics Journal","volume":"52 1","pages":"85-95"},"PeriodicalIF":0.3,"publicationDate":"2008-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81100203","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：