Biomedical subjects

William J Welsh

Publications and source records attributed to William J Welsh.

At least 19 recordsLinked to original sources

PubMed2005

DNA microarray data imputation and significance analysis of differential expression.

MOTIVATION: Significance analysis of differential expression in DNA microarray data is an important task. Much of the current research is focused on developing improved tests and software tools. The task is difficult not only owing to the high dimensionality of the data (number of genes), but also because of the often non-negligible presence of missing values. There is thus a great need to reliably impute these missing values prior to the statistical analyses. Many imputation methods have been developed for DNA microarray data, but their impact on statistical analyses has not been well studied. In this work we examine how missing values and their imputation affect significance analysis of differential expression. RESULTS: We develop a new imputation method (LinCmb) that is superior to the widely used methods in terms of normalized root mean squared error. Its estimates are the convex combinations of the estimates of existing methods. We find that LinCmb adapts to the structure of the data: If the data are heterogeneous or if there are few missing values, LinCmb puts more weight on local imputation methods; if the data are homogeneous or if there are many missing values, LinCmb puts more weight on global imputation methods. Thus, LinCmb is a useful tool to understand the merits of different imputation methods. We also demonstrate that missing values affect significance analysis. Two datasets, different amounts of missing values, different imputation methods, the standard t-test and the regularized t-test and ANOVA are employed in the simulations. We conclude that good imputation alleviates the impact of missing values and should be an integral part of microarray data analysis. The most competitive methods are LinCmb, GMC and BPCA. Popular imputation schemes such as SVD, row mean, and KNN all exhibit high variance and poor performance. The regularized t-test is less affected by missing values than the standard t-test. AVAILABILITY: Matlab code is available on request from the authors.

Algorithms↗

PubMed2005

Rapid assessment of contact-dependent secondary structure propensity: relevance to amyloidogenic sequences.

We have previously demonstrated that calculation of contact-dependent secondary structure propensity (CSSP) is highly sensitive in detecting non-native beta-strand propensities in the core sequences of known amyloidogenic proteins. Here we describe a CSSP method based on an artificial neural network that rapidly and accurately quantifies the influence of tertiary contacts (TCs) on secondary structure propensity in local regions of protein sequences. The present method exhibited 72% accuracy in predicting the alternate secondary structure adopted by chameleon sequences located in highly disparate TC regions. Analysis of 1930 nonhomologous protein domains reveals that the alpha-helix and the beta-strand largely share the same sequence context, and that tertiary context is a major determinant of the native conformation. Conversely, it appears that the propensity of random coils for either the alpha-helix or the beta-strand is largely invariant to tertiary effects. The present CSSP method successfully reproduced the amyloidogenic character observed in local regions of the human islet amyloid polypeptide (hIAPP). Furthermore, CSSP profiles were strongly correlated (r = 0.76) with the observed mutational effects on the aggregation rate of acylphosphatase. Taken together, these results provide compelling evidence in support of the present CSSP approach as a sensitive probe useful for analysis of full-length proteins and for detection of core sequences that may trigger amyloid fibril formation. The combined speed and simplicity of the CSSP method lends itself to proteome-wide analysis of the amyloidogenic nature of common proteins.

Algorithms↗

PubMed2005

Prediction of the orientations of adsorbed protein using an empirical energy function with implicit solvation.

When simulating protein adsorption behavior, decisions must first be made regarding how the protein should be oriented on the surface. To address this problem, we have developed a molecular simulation program that combines an empirical adsorption free energy function with an efficient configurational search method to calculate orientation-dependent adsorption free energies between proteins and functionalized surfaces. The configuration space is searched systematically using a quaternion rotation technique, and the adsorption free energy is evaluated using an empirical energy function with an efficient grid-based calculational method. In this paper, the developed method is applied to analyze the preferred orientations of a model protein, lysozyme, on various functionalized alkanethiol self-assembled monolayer (SAM) surfaces by the generation of contour graphs that relate adsorption free energy to adsorbed orientation, and the results are compared with experimental observations. As anticipated, the adsorbed orientation of lysozyme is predicted to be dependent on the discrete organization of the functional groups presented by the surface. Lysozyme, which is a positively charged protein, is predicted to adsorb on its 'side' on both hydrophobic and negatively charged surfaces. On surfaces with discrete positively charged sites, attractive interaction energies can also be obtained due to the presence of discrete local negative charges present on the lysozyme surface. In this case, 'end-on' orientations are preferred. Additionally, SAM surface models with mixed functionality suggest that the interactions between lysozyme and surfaces could be greatly enhanced if individual surface functional groups are able to access the catalytic cleft region of lysozyme, similar to ligand-receptor interactions. The contour graphs generated by this method can be used to identify low-energy orientations that can then be used as starting points for further simulations to investigate conformational changes induced in protein structure following initial adsorption.

Adsorption↗

PubMed2005

Inhibition of human liver catechol-O-methyltransferase by tea catechins and their metabolites: structure-activity relationship and molecular-modeling studies.

(-)-Epigallocatechin-3-gallate (EGCG) is the major polyphenol present in green tea. We previously demonstrated that EGCG was both a substrate and potent inhibitor of human liver cytosolic catechol-O-methyltransferease (COMT). We now report the structure-activity relationship for the inhibition of COMT-catalyzed O-methylation of catecholestrogens in human liver cytosol by tea catechins and some of their metabolites. The most potent inhibitors were catechins with a galloyl-type D-ring, including EGCG (IC(50)=0.07 microM), 4''-O-methyl-EGCG (IC(50)=0.10 microM), 4',4''-di-O-methyl-EGCG (4',4''-DiMeEGCG) (IC(50)=0.15 microM), and (-)-epicatechin-3-gallate (ECG) (IC(50)=0.20 microM). Catechins without the D-ring showed two to three orders of magnitude less inhibitory potency. Enzyme kinetic analyses revealed that EGCG behaved as a mixed inhibitor, whereas 4',4''-di-O-methyl-EGCG exhibited competitive kinetics for the S-adenosylmethionine (SAM), and noncompetitive kinetics for the catechol binding site. These compounds may represent a new type of COMT inhibitor. In silico molecular-modeling studies using a homology model of human COMT were conducted to aid in the understanding the catalytic and inhibitory mechanisms. Either D-ring or B-ring of EGCG could be accommodated to the substrate binding pocket of human COMT. However, the close proximity (2.6A) of 4''-OH to the critical residue Lys144, the higher acidity of the hydroxyl groups of the D-ring, and the hydrophobic interactions between the D-ring and residues in the binding pocket greatly facilitated the interaction of the D-ring with the enzyme, and resulted in increased inhibitory potency. These results provide mechanistic insight into the inhibition of COMT by commonly consumed tea catechins.

Catechin↗

PubMed2005

Benzotropolone inhibitors of estradiol methylation: kinetics and in silico modeling studies.

Natural and synthetic benzotropolone compounds were assessed in vitro for their ability to inhibit hydroxyestradiol methylation by catechol-O-methyltransferase (COMT). The compounds were also modeled in silico with a homology model of human COMT. Purpurogallin (1), purpurogallin carboxylic acid (2), and theaflavin-3,3'-digallate (6) were the most potent inhibitors of 2-hydroxy and 4-hydroxyestradiol methylation (IC(50) 0.22-0.50microM). Compounds 1 and 6 decreased the V(max) and increased the K(m) of COMT, indicating a mixed-type inhibition. Compounds 1 and 2 bound to COMT by inserting the six-membered ring of the benzotropolone into the active site. Decreased acidity of the hydroxyl groups on this ring or increased bulkiness reduced potency. Compound 6 bound by inserting the galloyl ester into the active site, which allowed the compound to overcome increased bulkiness and resulted in restored potency. Further studies are needed to determine the impact in vivo of COMT inhibition by these compounds.

Catechol O-Methyltransferase Inhibitors↗

PubMed2005

Improved method for predicting beta-turn using support vector machine.

MOTIVATION: Numerous methods for predicting beta-turns in proteins have been developed based on various computational schemes. Here, we introduce a new method of beta-turn prediction that uses the support vector machine (SVM) algorithm together with predicted secondary structure information. Various parameters from the SVM have been adjusted to achieve optimal prediction performance. RESULTS: The SVM method achieved excellent performance as measured by the Matthews correlation coefficient (MCC = 0.45) using a 7-fold cross validation on a database of 426 non-homologous protein chains. To our best knowledge, this MCC value is the highest achieved so far for predicting beta-turn. The overall prediction accuracy Qtotal was 77.3%, which is the best among the existing prediction methods. Among its unique attractive features, the present SVM method avoids overtraining and compresses information and provides a predicted reliability index.

Algorithms↗

PubMed2005

3D-QSAR comparative molecular field analysis on opioid receptor antagonists: pooling data from different studies.

Three-dimensional quantitative structure-activity relationship (3D-QSAR) models were constructed using comparative molecular field analysis (CoMFA) on a series of opioid receptor antagonists. To obtain statistically significant and robust CoMFA models, a sizable data set of naltrindole and naltrexone analogues was assembled by pooling biological and structural data from independent studies. A process of "leave one data set out", similar to the traditional "leave one out" cross-validation procedure employed in partial least squares (PLS) analysis, was utilized to study the feasibility of pooling data in the present case. These studies indicate that our approach yields statistically significant and highly predictive CoMFA models from the pooled data set of delta, mu, and kappa opioid receptor antagonists. All models showed excellent internal predictability and self-consistency: q(2) = 0.69/r(2) = 0.91 (delta), q(2) = 0.67/r(2) = 0.92 (mu), and q(2) = 0.60/r(2) = 0.96 (kappa). The CoMFA models were further validated using two separate test sets: one test set was selected randomly from the pooled data set, while the other test set was retrieved from other published sources. The overall excellent agreement between CoMFA-predicted and experimental binding affinities for a structurally diverse array of ligands across all three opioid receptor subtypes gives testimony to the superb predictive power of these models. CoMFA field analysis demonstrated that the variations in binding affinity of opioid antagonists are dominated by steric rather than electrostatic interactions with the three opioid receptor binding sites. The CoMFA steric-electrostatic contour maps corresponding to the delta, mu, and kappa opioid receptor subtypes reflected the characteristic similarities and differences in the familiar "message-address" concept of opioid receptor ligands. Structural modifications to increase selectivity for the delta over mu and kappa opioid receptors have been predicted on the basis of the CoMFA contour maps. The structure-activity relationships (SARs) together with the CoMFA models should find utility for the rational design of subtype-selective opioid receptor antagonists.

Least-Squares Analysis↗

PubMed2005

Elucidation of the Na+, K+-ATPase digitalis binding site.

Despite controversy over their use and the potential for toxic side effects, cardiac glycosides have remained an important clinical component for the treatment for congestive heart failure (CHF) and supraventricular arrhythmias since the effects of Digitalis purpurea were first described in 1785. While there is a wealth of information available with regard to the effects of these drugs on their pharmacological receptor, the Na(+), K(+)-ATPase, the exact molecular mechanism of digitalis binding and inhibition of the enzyme has remained elusive. In particular, the absence of structural knowledge about Na(+), K(+)-ATPase has thwarted the development of improved therapeutic agents with larger therapeutic indices via rational drug design approaches. Here, we propose a binding mode for digoxin and several analogues to the Na(+), K(+)-ATPase. A 3D-structural model of the extracellular loop regions of the catalytic alpha1-subunit of the digitalis-sensitive sheep Na(+), K(+)-ATPase was constructed from the crystal structure of an E(1)Ca(2+) conformation of the SERCA1a and a consensus orientation for digitalis binding was inferred from the in silico docking of a series of steroid-based cardiotonic compounds. Analyses of species-specific enzyme affinities for ouabain were also used to validate the model and, for the first time, propose a detailed model of the digitalis binding site.

Amino Acid Sequence↗

PubMed2005

3D-QSAR comparative molecular field analysis on delta opioid receptor agonist SNC80 and its analogs.

Three-dimensional quantitative structure-activity relationship (3D-QSAR) models were constructed using comparative molecular field analysis (CoMFA) for a series of delta opioid receptor agonists: SNC80 analogs. Quantum chemical calculations on SNC80 show that protonation is preferred at the basic N4 atom over the alternative N1 atom, accordingly N4 protonation may contribute significantly to ligand-receptor interactions under physiologically relevant conditions. Statistically significant and predictive CoMFA models were achieved by pooling biological data from independent published sources, including compounds with both alphaR and alphaS benzylic configurations. Improved CoMFA models were obtained when the compounds were considered as N4-protonated species rather than neutral compounds. The influence of various atomic partial-charge formalisms, alignment schemes and additional molecular descriptors was evaluated in order to produce the highest quality models. In addition, separate CoMFA models were generated for compounds with only the alphaR benzylic configuration. These CoMFA models showed excellent internal predictability and consistency, and external validation using test-set compounds yielded predicted pIC50 values within 1log unit of the corresponding experimentally measured values. Key insights into the structure-activity relationship derived from the CoMFA analysis concur with experimentally observed data, thus the CoMFA models presented here find utility for predicting the binding affinity, and guiding the design, of novel SNC80 analogs and related delta opioid receptor agonists.

Animals↗

PubMed2005

Structural model of the Plasmodium CDK, Pfmrk, a novel target for malaria therapeutics.

Malaria, with 300-500 million clinical cases resulting in 1-3 million fatalities a year, is one of the most deadly tropical diseases. As current antimalarial therapeutics become increasingly ineffective due to parasitic resistance, there exists an urgent need to develop and pursue new therapeutic strategies. Recent genome sequencing and molecular cloning projects have identified several enzymes from Plasmodium (P.) falciparum that may represent novel drug targets, including a family of proteins that are homologous to the mammalian cyclin-dependent kinases (CDKs). CDKs are essential for the control of the mammalian cell cycle and, based on the conservation of the CDKs across species, the plasmodial CDKs are expected to play a crucial role in parasitic growth. Here we present a 3D structural model of Pfmrk, a putative human CDK activating kinase (CAK) homolog in P. falciparum. Notable features of the present structural model include: (1) parameterization of the Mg2+ hexacoordination system using ab initio quantum chemical calculations to accurately represent the ATP-kinase interaction; and (2) comparison between the docking scores and measured binding affinities for a series of oxindole-based Pfmrk inhibitors of known activity. Detailed analysis of inhibitor-Pfmrk binding interactions enabled us to identify specific residues (viz. Met66, Met75, Met91, Met94 and Phe143) within the Pfmrk binding pocket that may play an important role in inhibitor binding affinity and selectivity. The availability of this Pfmrk structural model, together with insights gained from analysis of ligand-receptor interactions, should promote the rational design of potent and selective Pfmrk inhibitors as antimalarial therapeutics.

Amino Acid Sequence↗

PubMed2005

Requirement of activated Cdc42-associated kinase for survival of v-Ras-transformed mammalian cells.

Activated Cdc42-associated kinase (ACK) has been shown to be an important effector molecule for the small GTPase Cdc42. We have shown previously an essential role for Cdc42 in the transduction of Ras signals for the transformation of mammalian cells. In this report, we show that the ACK-1 isoform of ACK plays a critical role in transducing Ras-Cdc42 signals in the NIH 3T3 cells. Overexpression of a dominant-negative (K214R) mutant of ACK-1 inhibits Ras-induced up-regulation of c-fos and inhibits the growth of v-Ras-transformed NIH 3T3 cells. Using small interfering RNA, we knocked down the expression of ACK-1 in both v-Ha-Ras-transformed and parental NIH 3T3 cells and found that down-regulation of ACK-1 inhibited cell growth by inducing apoptosis only in v-Ha-Ras-transformed but not parental NIH 3T3 cells. In addition, we studied the effect of several tyrosine kinase inhibitors and found that PD158780 inhibits the kinase activity of ACK-1 in vitro. We also found that PD158780 inhibits the growth of v-Ha-Ras-transformed NIH 3T3 cells. Taken together, our results suggest that ACK-1 kinase plays an important role in the survival of v-Ha-Ras-transformed cells, suggesting that ACK-1 is a novel target for therapies directed at Ras-induced cancer.

Animals↗

PubMed2005

Rational inhibitor design and iterative screening in the identification of selective plasmodial cyclin dependent kinase inhibitors.

New chemical classes of compounds must be introduced into the malaria drug development pipeline in an effort to develop new chemotherapy options for the fight against malaria. In this review we describe an iterative approach designed to identify potent inhibitors of a kinase family that collectively functions as key regulators of the cell cycle. Cyclin-dependent protein kinases (CDKs) are attractive drug targets in numerous diseases and, most recently, they have become the focus of rational drug design programs for the development of new antimalarial agents. Our approach uses experimental and virtual screening methodologies to identify and refine chemical inhibitors and increase the success rate of discovering potent and selective inhibitors. The active pockets of the plasmodial CDKs are unique in terms of size, shape and amino acid composition compared with those of the mammalian orthologues. These differences exemplified through the use of screening assays, molecular modeling, and crystallography can be exploited for inhibitor design. To date, several classes of compounds including quinolines and oxindoles have been identified as selective inhibitors of the plasmodial CDK7 homologue, Pfmrk. From these initial studies and through the iterative rational drug design process, more potent, selective, and most importantly, chemically unique compound classes have been identified as effective inhibitors of the plasmodial CDKs and the malarial parasite.

Amino Acid Sequence↗

PubMed2004

Identification of possible kinetically significant anion-binding sites in human serum transferrin using molecular modeling strategies.

Certain anions have been shown experimentally to influence the rate of iron release from human serum transferrin (HST), implying the existence of one or more allosteric kinetically significant anion-binding (KISAB) sites on or near the surface of the protein. A rank-ordered selection of potential HST KISAB sites has been obtained using a novel three-stage molecular modeling strategy. The crystal structure of HST (1A8E.pdb) was first subjected to a heuristic analysis, in which positively charged and hydrogen-bonding residues on or near the surface of the protein were identified. In this stage, a preliminary electrostatic potential map was also calculated, yielding six preliminary sites. Next, energy-grid calculations were conducted in order to identify anion-protein interaction energy minima, which resulted in the inclusion of three additional sites. Finally, three anions already shown experimentally to demonstrate varied effects on HST iron-release kinetics were placed at each potential site; molecular dynamics and molecular mechanics calculations were performed in order to elucidate the hydrogen-bonding environment around each anion of the protein as well as to calculate anion-protein-binding energies.

Anions↗

PubMed2004

Gaussian mixture clustering and imputation of microarray data.

MOTIVATION: In microarray experiments, missing entries arise from blemishes on the chips. In large-scale studies, virtually every chip contains some missing entries and more than 90% of the genes are affected. Many analysis methods require a full set of data. Either those genes with missing entries are excluded, or the missing entries are filled with estimates prior to the analyses. This study compares methods of missing value estimation. RESULTS: Two evaluation metrics of imputation accuracy are employed. First, the root mean squared error measures the difference between the true values and the imputed values. Second, the number of mis-clustered genes measures the difference between clustering with true values and that with imputed values; it examines the bias introduced by imputation to clustering. The Gaussian mixture clustering with model averaging imputation is superior to all other imputation methods, according to both evaluation metrics, on both time-series (correlated) and non-time series (uncorrelated) data sets.

Algorithms↗

PubMed2004

Characteristics of the Plasmodium falciparum PK5 ATP-binding site: implications for the design of novel antimalarial agents.

Increasing worldwide resistance of Plasmodium falciparum (P. falciparum) to traditional chemotherapy strategies such as chloroquine and mefloquine demonstrates the urgent need for the discovery of novel chemotherapeutic agents in the fight against malaria. The recent discovery of P. falciparum Protein Kinase 5 (PfPK5) invites the possibility of selectively targeting the life cycle of P. falciparum in order to prevent cerebral malaria. PfPK5 bears a high degree of sequence identity (>58%) to a structurally conserved family of mammalian kinases known as the cyclin-dependent kinases (CDKs). The CDKs are the key regulatory elements governing the ordered progression of the mammalian cell cycle. With numerous X-ray crystal structures of CDK2 to provide a structural template, here we present a three-dimensional structural model of PfPK5 constructed using computer-based homology modeling techniques. Our model was used to compare the ATP binding site of PfPK5 with that of the mammalian kinase CDK2. Furthermore, kinase-ligand interactions of PfPK5 with known inhibitors were investigated and compared to available crystal structures of CDK2 with inhibitors bound. The focus of the study is to identify similarities and differences between the ATP binding sites of the two kinases that can be exploited for future rational drug design.

Adenosine Triphosphate↗

PubMed2004

Detecting hidden sequence propensity for amyloid fibril formation.

The preponderance of evidence implicates protein misfolding in many unrelated human diseases. In all cases, normal correctly folded proteins transform from their proper native structure into an abnormal beta-rich structure known as amyloid fibril. Here we introduce a computational algorithm to detect nonnative (hidden) sequence propensity for amyloid fibril formation. Analyzing sequence-structure relationships in terms of tertiary contact (TC), we find that the hidden beta-strand propensity of a query local sequence can be quantitatively estimated from the secondary structure preferences of template sequences of known secondary structure found in regions of high TC. The present method correctly pinpoints the minimal peptide fragment shown experimentally as the likely local mediator of amyloid fibril formation in beta-amyloid peptide, islet amyloid polypeptide (hIAPP), alpha-synuclein, and human acetylcholinesterase (AChE). It also found previously unrecognized beta-strand propensities in the prototypical helical protein myoglobin that has been reported as amyloidogenic. Analysis of 2358 nonhomologous protein domains provides compelling evidence that most proteins contain sequences with significant hidden beta-strand propensity. The present method may find utility in many medically relevant applications, such as the engineering of protein sequences and the discovery of therapeutic agents that specifically target these sequences for the prevention and treatment of amyloid diseases.

Acetylcholinesterase↗

PubMed2004

Highly chlorinated PCBs inhibit the human xenobiotic response mediated by the steroid and xenobiotic receptor (SXR).

Polychlorinated biphenyls (PCBs) are a family of persistent organic contaminants suspected to cause adverse effects in wildlife and humans. In rodents, PCBs bind to the aryl hydrocarbon (AhR) and pregnane X receptors (PXR) inducing the expression of catabolic cytochrome p450 enzymes of the CYP1A and 3A families. We found that certain highly chlorinated PCBs are potent activators of rodent PXR but antagonize its human ortholog, the steroid and xenobiotic receptor (SXR), inhibiting target gene induction. Thus, exposure to PCBs may blunt the human xenobiotic response, inhibiting the detoxification of steroids, bioactive dietary compounds, and xenobiotics normally mediated by SXR. The antagonistic PCBs are among the most stable and abundant in human tissues. These findings have important implications for understanding the biologic effects of PCB exposure and the use of animal models to predict the attendant risk.

Animals↗

PubMed2003

Shape signatures: a new approach to computer-aided ligand- and receptor-based drug design.

A unifying principle of rational drug design is the use of either shape similarity or complementarity to identify compounds expected to be active against a given target. Shape similarity is the underlying foundation of ligand-based methods, which seek compounds with structure similar to known actives, while shape complementarity is the basis of most receptor-based design, where the goal is to identify compounds complementary in shape to a given receptor. These approaches can be extended to include molecular descriptors in addition to shape, such as lipophilicity or electrostatic potential. Here we introduce a new technique, which we call shape signatures, for describing the shape of ligand molecules and of receptor sites. The method uses a technique akin to ray-tracing to explore the volume enclosed by a ligand molecule, or the volume exterior to the active site of a protein. Probability distributions are derived from the ray-trace, and can be based solely on the geometry of the reflecting ray, or may include joint dependence on properties, such as the molecular electrostatic potential, computed over the surface. Our shape signatures are just these probability distributions, stored as histograms. They converge rapidly with the length of the ray-trace, are independent of molecular orientation, and can be compared quickly using simple metrics. Shape signatures can be used to test for both shape similarity between compounds and for shape complementarity between compounds and receptors and thus can be applied to problems in both ligand- and receptor-based molecular design. We present results for comparisons between small molecules of biological interest and the NCI Database using shape signatures under two different metrics. Our results show that the method can reliably extract compounds of shape (and polarity) similar to the query molecules. We also present initial results for a receptor-based strategy using shape signatures, with application to the design of new inhibitors predicted to be active against HIV protease.

Binding Sites↗