Sergio Clementi

FLAP: GRID molecular interaction fields in virtual screening. validation using the DUD data set.

The performance of FLAP (Fingerprints for Ligands and Proteins) in virtual screening is assessed using a subset of the DUD (Directory of Useful Decoys) benchmarking data set containing 13 targets each with more than 15 different chemotype classes. A variety of ligand and receptor-based virtual screening approaches are examined, using combinations of individual templates 2D structures of known actives, a cocrystallized ligand, a receptor structure, or a cocrystallized ligand-biased receptor structure. We examine several data fusion approaches to combine the results of the individual virtual screens. In doing so, we show that excellent chemotype enrichment is achieved in both single target ligand-based and receptor-based approaches, of approximately 17-fold over random on average at a false positive rate of 1%. We also show that using as much starting knowledge as possible improves chemotype enrichment, and that data fusion using Pareto ranking is an effective method to do this giving up to 50% improvement in enrichment over the single methods. Finally we show that if inactivity or decoy data is incorporated, automatically training the scoring function in FLAP improves recovery still further, with almost 2-fold improvement over the enrichments shown by the single methods. The results clearly demonstrate the utility of FLAP for virtual screening when either a limited or wide range of prior knowledge is available.

Some applications of the partial least-squares method

Abstract The philosophy of the applications of the partial least-squares method in chemistry is outlined. The method is compared with multiple regression and illustrated by means of examples in various fields of chemistry.

Multivariate chemical mapping of pigments and binders in easel painting cross-sections by micro IR reflection spectroscopy

AbstractPaintings are composed of superimposed layers of inorganic and organic materials (pigments and binders). Knowledge of the stratigraphic sequence of these heterogeneous layers is fundamental for understanding the artist’s painting technique and for conservation issues. In this study, micro-IR mapping experiments in reflection mode have been carried out on cross-sections taken from simulations of ancient easel paintings. The objective was to locate both organic binders and inorganic pigments. Chemical maps have been re-constructed using the common approach based on the integration of specific infrared bands. However, owing to the complexity of painting materials, this approach is not always applicable when dealing with broad and superimposed spectral features and with reststrahlen or derivative-like bands resulting from acquisition in reflection mode. To overcome these limitations, principal-component analysis has been successfully used for the re-construction of the image, extracting the relevant information from the complex full spectral data sets and obtaining reliable chemical distributions of the stratigraphy materials. Different pigment–binder combinations have been evaluated in order to understand limitations and strengths of the approach. Finally, the method has been applied for stratigraphic characterization of a cross-section from a 17th century wooden sculpture identifying both the original paint layer and the several overpaintings constituting the complex stratigraphy. Figurea) OM image of an easel painting cross-section with the grid indicating the IR mapped area; b) t1 vs. t2 score plot including the color bar referred to the score values along each components; c) t2-score map; d) p1 and p2 profiles from PCA (labels: Gy-gypsum, L-lipid, I-amide I, II-amide II, U-ultramarine blue)

Application of principal components analysis to the evaluation and selection of eluent systems for the thin-layer chromatography of basic and neutral drugs

Abstract The R F values of 55 drugs in 40 eluent mixtures are reported. Principal component analysis of these data provides a four-significant-components model, which explains 92% of the total variance. This analysis, showing that the eluent mixtures cluster into different groups according to their information content, provides a reliable criterion for the choice of optimal eluents. Four eluent mixtures of [ethyl acetate—methanol—30% ammonia (85:10:5), cyclohexane—toluene—diethylamine (65:25:10), ethyl acetate—chloroform (50:50) and acetone with the plate dipped in potassium hydroxide solution), chosen on the basis of the above criterion and of the R F reproducibility, provide a two significant principal components model that can be used for the identification of unknown samples.

THE IMPORTANCE OF SECONDARY ANCHOR RESIDUE MOTIFS OF HLA CLASS I PROTEINS: A CHEMOMETRIC APPROACH

In this paper we report a chemometric approach to Quantitative Structure-Activity Relationship (QSAR) analysis applied to a study of the binding of peptides to Major Histocompatibility Complex (MHC) class I proteins. Peptides which possess the known primary anchor residue motif for HLA-B27 binding do not necessarily bind to HLA-B27 proteins. Secondary anchor residues are also involved, but it is not yet clear which amino acids are required or in which positions. A classic approach to this problem would be to synthesize multiple peptides each varying by a single amino acid from a starting peptide, and test them for their binding properties. Not only is this approach inefficient, but it is essentially unable to provide information about possible mutual interactions of amino acid residues in different positions. Using a statistical design to select the most informative compounds to use in the QSAR study, it was possible to analyse the effects on HLA-B27 peptide binding of different amino acids in four positions by means of only nine peptides. The relative binding activity of these peptides could then be modeled mathematically to provide information about the relative contribution of each of the four positions and to suggest a new peptide with high binding affinity. Our results demonstrate the usefulness of the chemometric strategy for studying peptides of interest in molecular immunology.

Clustering of aryl carbon-13 nuclear magnetic resonance substituent chemical shifts. A multivariate data analysis using principal components

A principal component analysis of the 13C substituent-induced chemical shifts of 82 monosubstituted benzenes shows that ca. 90% of the substituents belong to one of four clusters, acceptors, alkyls, donors, or halogens. This grouping is confirmed statistically. The extensions of the subclasses are not parallel. It is also shown that the predictive capability of the single-parameter models for each subclass is better than any multiparameter model applied on the whole data set. The observed grouping of substituents provides an explanation to the apparent correlation frequently found between 13C n.m.r. chemical shifts and dual substituent parameters. The ability of the statistical method to discover incorrect shift data is also illustrated.

Quantitative structure-property relationships study of azo dyes using partial least Squares analysis in latent variables (PLS)

Abstract The light, washing, alkali and perspiration fastness of a series of azo dyes are analysed as a function of physicochemical substituent descriptors by the PLS method. The analysis permits relation of the fastness properties to the structural features and the development of a model which can be used to predict the fastness of new dyes of the same series.

Principal components and partial least-squares analysis of the geochemistry of volcanic rocks from the aeolian archipelago

Abstract A data set from total chemical analysis of 53 volcanic rocks from five islands of the Aeolian Archipelago (11 major and 13 trace elements) was examined by the multivariate SIMCA and PLS data analysis programs, in an attempt to evaluate the geochemical potential of the methods. The results outline: (a) the optimal within-islands classification and discrimination variables; and (b) the relative information content of the blocks of major elements and trace elements. A measure of the common systematic information of these two blocks is also obtained. The study illustrates blockwise, related pattern recognition of the level-4 type.

Partial least-squares modelling of dye fastness to light

Abstract Several physicochemical properties ( 13 C-n.m.r., spectrophotometric, chromatographic, calorimetric) were measured for a series of 38 azo dyes in order to test the existence of relationships between one or more of these data sets and the fastness of the dyes on polyester fabric. The partial least-squares method was applied; it had previously proved useful in establishing structure/property relationships for the same dyes. It is shown that 13 C-n.m.r. data can be used successfully to predict dye fastness to light. These data are quickly and cheaply obtained with small amounts of compound.

Unusual site of electrophilic attack on indole and carbazole

Abstract Whereas trifluoroacetylation of pyrrole in 1,2-dichloroethane gives exclusively C-substitution and carbazole gives exclusively N-substitution, trifluroacetylation of indole gives a mixture of N- and C-derivatives. This behaviour is discussed in terms of different loss of resonance energy in the transition state leading to N-substitution. Formation of N-trifluoroacetyl-2-(2-indolyl)indoline in trifluoroacetylation of indole is also discussed.