Results 1  10
of
28
Variable Neighborhood Search for Extremal Graphs 6. Analyzing Bounds for the Connectivity Index
, 2000
"... Recently, Araujo and De la Pe~na [1] gave bounds for the connectivity index of chemical trees as a function of this index for general trees and the ramification index of trees. They also gave bounds for the connectivity index of chemical graphs as a function of this index for maximal subgraphs which ..."
Abstract

Cited by 17 (5 self)
 Add to MetaCart
Recently, Araujo and De la Pe~na [1] gave bounds for the connectivity index of chemical trees as a function of this index for general trees and the ramification index of trees. They also gave bounds for the connectivity index of chemical graphs as a function of this index for maximal subgraphs which are trees and the cyclomatic number of the graphs. The ramification index of a tree is first shown to be equal to the number of pending vertices minus 2. Then, in view of extremal graphs obtained with the system AutoGraphiX, all bounds of Araujo and De la Pe\~na [1] are improved, yielding tight bounds, and in one case corrected. Moreover, chemical trees of given order and number of pending vertices with minimum and with maximum connectivity index are characterized.
Effect of molecular descriptor feature selection in support vector machine classification of pharmacokinetic and toxicological properties of chemical agents
 J. Chem. Inf. Comput. Sci
, 2004
"... Statisticallearning methods have been developed for facilitating the prediction of pharmacokinetic and toxicological properties of chemical agents. These methods employ a variety of molecular descriptors to characterize structural and physicochemical properties of molecules. Some of these descripto ..."
Abstract

Cited by 11 (3 self)
 Add to MetaCart
Statisticallearning methods have been developed for facilitating the prediction of pharmacokinetic and toxicological properties of chemical agents. These methods employ a variety of molecular descriptors to characterize structural and physicochemical properties of molecules. Some of these descriptors are specifically designed for the study of a particular type of properties or agents, and their use for other properties or agents might generate noise and affect the prediction accuracy of a statistical learning system. This work examines to what extent the reduction of this noise can improve the prediction accuracy of a statistical learning system. A feature selection method, recursive feature elimination (RFE), is used to automatically select molecular descriptors for support vector machines (SVM) prediction of Pglycoprotein substrates (Pgp), human intestinal absorption of molecules (HIA), and agents that cause torsades de pointes (TdP), a rare but serious side effect. RFE significantly reduces the number of descriptors for each of these properties thereby increasing the computational speed for their classification. The SVM prediction accuracies of Pgp and HIA are substantially increased and that of TdP remains unchanged by RFE. These prediction accuracies are comparable to those of earlier studies derived from a selective set of descriptors. Our study suggests that molecular feature selection is useful for improving the speed and, in some cases, the accuracy of statistical learning methods for the prediction of pharmacokinetic and toxicological properties of chemical agents.
Introducing Spectral Structure Activity Relationship (SSAR) Analysis. Application to Ecotoxicology
, 2007
"... Abstract: A novel quantitative structureactivity (property) relationship model, namely SpectralSAR, is presented in an exclusive algebraic way replacing the oldfashioned multiregression one. The actual SSAR method interprets structural descriptors as vectors in a generic data space that is furt ..."
Abstract

Cited by 10 (6 self)
 Add to MetaCart
Abstract: A novel quantitative structureactivity (property) relationship model, namely SpectralSAR, is presented in an exclusive algebraic way replacing the oldfashioned multiregression one. The actual SSAR method interprets structural descriptors as vectors in a generic data space that is further mapped into a full orthogonal space by means of the GramSchmidt algorithm. Then, by coordinated transformation between the data and orthogonal spaces, the SSAR equation is given under simple determinant form for any chemicalbiological interactions under study. While proving to give the same analytical equation and correlation results with standard multivariate statistics, the actual SSAR frame allows the introduction of the spectral norm as a valid substitute for the correlation factor, while also having the advantage to design the various related SAR models through the introduced “minimal spectral path ” rule. An application is given performing a complete SSAR analysis upon the Tetrahymena pyriformis ciliate species employing its reported ecotoxicity activities among relevant classes of xenobiotics. By representing the spectral norm of the endpoint models against the concerned structural coordinates, the obtained SSAR endpoints hierarchy scheme opens the perspective to further design the ecotoxicological test batteries with organisms from different species.
MOLecular Structure GENeration with MOLGEN, new features and future developments
 Fresenius J. Anal. Chem
, 1997
"... MOLGEN is a computer program system which is designed for generating molecular graphs fast, redundancy free and exhaustively. In the present paper we describe its basic features, new features of the current release MOLGEN 3.5, and future developments which provide considerable improvements and ex ..."
Abstract

Cited by 6 (4 self)
 Add to MetaCart
MOLGEN is a computer program system which is designed for generating molecular graphs fast, redundancy free and exhaustively. In the present paper we describe its basic features, new features of the current release MOLGEN 3.5, and future developments which provide considerable improvements and extensions. 1 Introduction MOLGEN [17] is a generator for molecular graphs (=connectivity isomers or constitutional formulae) allowing to generate all isomers that correspond to a given molecular formula and (optional) further conditions like prescribed and forbidden substructures, ring sizes etc. The input consists of ffl the empirical formula, together with ffl an optional list of macroatoms, which means prescribed substructures that must not overlap, ffl an optional goodlist, that consists of prescribed substructures which may overlap, ffl an optional badlist, containing forbidden substructures, ffl an optional interval for the minimal and maximal size of rings, ffl an optional num...
Prediction of TorsadeCausing Potential of Drugs by Support Vector Machine Approach
 Clinical Chemistry and Laboratory Medicine
, 2004
"... In an effort to facilitate drug discovery, computational methods for facilitating the prediction of various adverse drug reactions (ADRs) have been developed. So far, attention has not been sufficiently paid to the development of methods for the prediction of serious ADRs that occur less frequently. ..."
Abstract

Cited by 5 (3 self)
 Add to MetaCart
In an effort to facilitate drug discovery, computational methods for facilitating the prediction of various adverse drug reactions (ADRs) have been developed. So far, attention has not been sufficiently paid to the development of methods for the prediction of serious ADRs that occur less frequently. Some of these ADRs, such as torsade de pointes (TdP), are important issues in the approval of drugs for certain diseases. Thus there is a need to develop tools for facilitating the prediction of these ADRs. This work explores the use of a statistical learning method, support vector machine (SVM), for TdP prediction. TdP involves multiple mechanisms and SVM is a method suitable for such a problem. Our SVM classification system used a set of linear solvation energy relationship (LSER) descriptors and was optimized by leaveoneout cross validation procedure. Its prediction accuracy was evaluated by using an independent set of agents and by comparison
Classification of small molecules by two and threedimensional decomposition kernels
 BIOINFORMATICS
, 2007
"... ..."
Use of Statistical and Neural Net Methods in Predicting Toxicity of Chemicals: A Hierarchical QSAR Approach
 Predictive Toxicology of Chemicals: Experiences and Impacts of AI Tools
, 1999
"... A contemporary trend in computational toxicology is the prediction of toxicity endpoints and toxic modes of action of chemicals from parameters that can be calculated directly from their molecular structure. Topological, geometrical, substructural, and quantum chemical parameters fall into this ..."
Abstract

Cited by 3 (0 self)
 Add to MetaCart
A contemporary trend in computational toxicology is the prediction of toxicity endpoints and toxic modes of action of chemicals from parameters that can be calculated directly from their molecular structure. Topological, geometrical, substructural, and quantum chemical parameters fall into this category. We have been involved in the development of a new hierarchical quantitative structureactivity relationship (QSAR) approach in predicting physicochemical, biomedicinal and toxicological properties of various sets of chemicals. This approach uses increasingly more complex molecular descriptors for model building in a graduated manner. In this paper we will apply statistical and neural net methods in the development of QSAR models for predicting toxicity of chemicals using topostructural, topochemical, geometrical, and quantum chemical indices. The utility and limitations of the approach will be discussed. Introduction In 1998 the number of chemicals registered with the...
Trees of extremal connectivity index
 Discrete Appl. Math
"... The connectivity index wα(G) of a graph G is the sum of the weights (d(u)d(v)) α of all edges uv of G, where α is a real number (α � = 0), and d(u) denotes the degree of the vertex u. Let T be a tree with n vertices and k pendant vertices. In this paper, we give sharp lower and upper bounds for w1(T ..."
Abstract

Cited by 3 (1 self)
 Add to MetaCart
The connectivity index wα(G) of a graph G is the sum of the weights (d(u)d(v)) α of all edges uv of G, where α is a real number (α � = 0), and d(u) denotes the degree of the vertex u. Let T be a tree with n vertices and k pendant vertices. In this paper, we give sharp lower and upper bounds for w1(T). Also, for −1 ≤ α < 0, we give a sharp lower bound and a upper bound for wα(T).
Prediction of Pglycoprotein substrates by a support vector machine approach
 J Chem Inf Comput Sci
, 2004
"... Pglycoproteins (Pgp) actively transport a wide variety of chemicals out of cells and function as drug efflux pumps that mediate multidrug resistance and limit the efficacy of many drugs. Methods for facilitating early elimination of potential Pgp substrates are useful for facilitating new drug di ..."
Abstract

Cited by 3 (2 self)
 Add to MetaCart
Pglycoproteins (Pgp) actively transport a wide variety of chemicals out of cells and function as drug efflux pumps that mediate multidrug resistance and limit the efficacy of many drugs. Methods for facilitating early elimination of potential Pgp substrates are useful for facilitating new drug discovery. A computational ensemble pharmacophore model has recently been used for the prediction of Pgp substrates with a promising accuracy of 63%. It is desirable to extend the prediction range beyond compounds covered by the known pharmacophore models. For such a purpose, a machine learning method, support vector machine (SVM), was explored for the prediction of Pgp substrates. A set of 201 chemical compounds, including 116 substrates and 85 nonsubstrates of Pgp, was used to train and test a SVM classification system. This SVM system gave a prediction accuracy of at least 81.2 % for Pgp substrates based on two different evaluation methods, which is substantially improved against that obtained from the multiplepharmacophore model. The prediction accuracy for nonsubstrates of Pgp is 79.2 % using 5fold crossvalidation. These accuracies are slightly better than those obtained from other statistical classification methods, including knearest neighbor (kNN), probabilistic neural networks (PNN), and C4.5 decision tree, that use the same sets of data and molecular descriptors. Our study indicates the potential of SVM in facilitating the prediction of Pgp substrates.
On the Randić Index ∗
"... The Randić index of an organic molecule whose molecular graph is G is defined 1 − as the sum of (d(u)d(v)) 2 over all pairs of adjacent vertices of G, where d(u) is the degree of the vertex u in G. In [2], Delorme et al gave a bestpossible lower bound on the Randić index of a trianglefree graph G ..."
Abstract

Cited by 3 (0 self)
 Add to MetaCart
The Randić index of an organic molecule whose molecular graph is G is defined 1 − as the sum of (d(u)d(v)) 2 over all pairs of adjacent vertices of G, where d(u) is the degree of the vertex u in G. In [2], Delorme et al gave a bestpossible lower bound on the Randić index of a trianglefree graph G with given minimum degree δ(G). In the paper, we first point out a careless mistake in the proof of their result (Theorem 2 of [2]), and then we will show that the result holds when δ(G) ≥ 2.