Multidimensional activity data

A spectral mapping technique, based on principal component analysis, has been developed for the two-dimensional interpretation of multidimensional activity data. It was successfully applied to characterize the activity profiles of drugs according to their effects in different pharmacological test models (e.g. Figure 45) [802-807]. 

Reverse engineering has been successfully applied to relatively simple biomolecular systems. Using a combination of cross-correlation analysis and multidimensional scaling, the glycolytic pathway was reconstructed from metabolite activity data from an in vitro enzymatic reactor system [21]. A complete spatio-temporal model of developmental gene expression in Drosophila was constructed for a small gene set based on models of differential equations and protein expression data [22],... 

ACTIVE DATA REPOSITORY FOR MULTIDIMENSIONAL DATA MANAGEMENT FOR TRANSLATIONAL LIFE SCIENCES RESEARCH... 

Large data sets such as screening data or results obtained by combinatorial experiments are made up of a large number of data records. Hence a data record may represent a chemical reaction or substance, for example its corresponding variables will define the corresponding reaction conditions or biological activities. Depending on the dimensionality or data type of the information, one-, two-, multidimensional, or specific data types can be identified. 

Discriminant emalysis is a supervised learning technique which uses classified dependent data. Here, the dependent data (y values) are not on a continuous scale but are divided into distinct classes. There are often just two classes (e.g. active/inactive soluble/not soluble yes/no), but more than two is also possible (e.g. high/medium/low 1/2/3/4). The simplest situation involves two variables and two classes, and the aim is to find a straight line that best separates the data into its classes (Figure 12.37). With more than two variables, the line becomes a hyperplane in the multidimensional variable space. Discriminant analysis is characterised by a discriminant function, which in the particular case of hnear discriminant analysis (the most popular variant) is written as a linear combination of the independent variables ... 

Limited protein stability often hampers successful structure elucidation by X-ray crystallography and/or NMR spectroscopy. Relaxation properties are usually improved at elevated temperatures, and multidimensional NMR experiments require sample lifetimes to extend over several days to weeks in order to acquire all the necessary data. In addition, the activity of contaminating proteases that are sometimes present in purified samples can be significant at the experimental temperatures. Therefore, the stability of a target protein can be a concern, in particular for expensive isotope-labeled proteins. 

On the other hand, factor analysis involves other manipulations of the eigen vectors and aims to gain insight into the structure of a multidimensional data set. The use of this technique was first proposed in biological structure-activity relationship (i. e., SAR) and illustrated with an analysis of the activities of 21 di-phenylaminopropanol derivatives in 11 biological tests [116-119, 289]. This method has been more commonly used to determine the intrinsic dimensionality of certain experimentally determined chemical properties which are the number of fundamental factors required to account for the variance. One of the best FA techniques is the Q-mode, which is based on grouping a multivariate data set based on the data structure defined by the similarity between samples [1, 313-316]. It is devoted exclusively to the interpretation of the inter-object relationships in a data set, rather than to the inter-variable (or covariance) relationships explored with R-mode factor analysis. The measure of similarity used is the cosine theta matrix, i. e., the matrix whose elements are the cosine of the angles between all sample pairs [1,313-316]. 

The paradigm shift from critical activities from later drug development to earlier discovery phases some years ago has effectively led to a change in lead optimization and added a new dimension of complexity, while it is envisioned that from a multidimensional, data-driven process more suitable candidates in accord with the therapeutic target product profiles may emerge for the treatment of currently unmet medical needs. 

This chapter is devoted to tunneling effects observed in vibration-rotation spectra of isolated molecules and dimers. The relative simplicity of these systems permits one to treat them in terms of multidimensional PES s and even to construct these PES s by using the spectroscopic data. Modern experimental techniques permit the study of these simple systems at superlow temperatures where tunneling prevails over thermal activation. The presence of large-amplitude anharmonic motions in these systems, associated with weak (e.g., van der Waals) forces, requires the full power of quantitative multidimensional tunneling theory. 

The need to maintain this data induces complex multidimensional information systems that few firms will even attempt to tackle. Clearly, this complex multidimensionality of source data exerts a profound effect on viewing and interpreting biological results correctly and efficiently. With natural products, it becomes very obvious that the traditional structure-activity relationship (SAR) rendition of results, so important to biochemists, becomes pointless due to the lack of a structure and the addition of multidimensional relationships between and within studied samples. 

It is clear that for an unsymmetrical data matrix that contains more variables (the field descriptors at each point of the grid for each probe used for calculation) than observables (the biological activity values), classical correlation analysis as multilinear regression analysis would fail. All 3D QSAR methods benefit from the development of PLS analysis, a statistical technique that aims to find the multidimensional direction in the X space that explains the maximum multidimensional variance direction in the F space. PLS is related to principal component analysis (PCA)." ° However, instead of finding the hyperplanes of maximum variance, it finds a linear model describing some predicted variables in terms of other observable variables and therefore can be used directly for prediction. Complexity reduction and data... 

AN ACTIVE, AS OPPOSED TO PASSIVE, MULTIDIMENSIONAL DATA REPOSITORY FOR BIOMEDICAL INFORMATICS... 

As shown in this review, the complexity of pharmacophores can range from very simple objects (two- or three-point pharmacophores) to more sophisticated objects by the addition of more pharmacophoric features, different types of geometric constraints, shape, or excluded regions information. 2D (substructure) as well as ID (relational data) information can also be added to a 3D pharmacophore. The nature of the pharmacophoric points (feature vs. substructure) will directly affect the overall performance of a database search. In general, an overspecification of the pharmacophoric points will result in hit lists with limited structural diversity. However, the use of pharmacophores is an efficient procedure since it eliminates quickly molecules that do not possess the required features. Unfortunately, all the retrieved hits are not always active as expected since the presence of the pharmacophoric groups is only one of the multiple components that account for the activity of a molecule. Other properties (physicochemical, ADME, and toxicological properties) are other components of the multidimensional approach that is used to turn a hit into a drug. 

We believe that data mining techniques will find utility in the future in determining accurate outputs from complex piezoelectric sensors yet to be developed. In the future, ever more complex biosensors and chemical sensors will be created on piezoelectric platforms. The accurate analyses of complex multidimensional inputs from such sensors may critically depend upon the use of machine learning algorithms. Such algorithms will learn to identify characteristic non-linear features of the inputs and associate them accurately with particular outputs (classification activity), such as an analyte concentration, that are then reported to the end-user. 

---