Mapping chemical space

The gas phase limitation imposed by the very nature of the GC-MS experiment limits its applicability to chemical space. The Venn diagram in Fig. 19.4 illustrates an approximate mapping of applicability of the combined chromatography-mass spectrometry techniques discussed here. Using only relative molecular mass and polarity as two dimensions that can map chemical space, the figure indicates that the applicability of GC-MS is limited to those relatively small, thermally stable, nonpolar compounds that can be readily volatized (or made to volatilize by derivatization [26]) to pass through a gas chromatographic column. 

Mapping Chemical Space by Self-organizing Maps A Pharmacophore Road Map... 

Rg. 6-10 Mapping chemical space [76]. Principle component models of chemical space are shown for 480 small molecules analyzed using 24 computed molecular descriptors and 60 measured phenotypic descriptors derived from a cell-based assay of cell proliferation. By considering the elements of S as coordinates, small molecules can be modeled as vectors,... 

Fig. 6 -12 Mapping chemical space using multidimensional phenotypic descriptors. Phenotypic data from multiple assays are arranged in a chemical-genetic data array and computational methods are used to select small molecules for further...

Wong, S.L. Schreiber, Mapping chemical space using molecular descriptors and chemical genetics deacetylase inhibitors, Comb. Chem. 

A data set can be split into a training set and a test set randomly or according to a specific rule. The 1293 compounds were divided into a training set of 741 compounds and a test set ot 552 compounds, based on their distribution in a K.NN map. From each occupied neuron, one compound was selected and taken into the training set, and the other compounds were put into the test set. This selection ensured that both the training set and the test set contained as much information as possible, and covered the chemical space as widely as possible. 

Oprea, T., Zamora, I., Ungell, A.-L., Pharmacokinetically based mapping device for chemical space navigation,... 

How is dimension reduction of chemical spaces achieved There are a number of different concepts and mathematical procedures to reduce the dimensionality of descriptor spaces with respect to a molecular dataset under investigation. These techniques include, for example, linear mapping, multidimensional scaling, factor analysis, or principal component analysis (PCA), as reviewed in ref. 8. Essentially, these techniques either try to identify those descriptors among the initially chosen ones that are most important to capture the chemical information encoded in a molecular dataset or, alternatively, attempt to construct new variables from original descriptor contributions. A representative example will be discussed below in more detail. 

An alternative to dimension reduction is the use of composite and uncorrelated descriptors that are suitable for the design of information-rich yet low-dimensional chemical spaces. An elegant example is presented by the popular BCUT (Burden-CAS-University of Texas) descriptors (Pearlman and Smith 1998). BCUTs are a set of uncorrelated descriptors that combine information about molecular connectivity, inter-molecular distances, and other molecular properties. BCUT spaces used for many applications are typically only six-dimensional and can frequently be further reduced to 3D representations for visualization purposes by identifying those BCUT axes around which most compounds map. 

Computationally, a structure space (represented as a set of two-dimensional graphs of molecule structures) is mapped to property (or chemical) space (Kp) (for example, Figure 2), where each point is a vector of values of each of p variables,... 

In addition, there have also been a number of efforts to find more efficient ways to categorise and enumerate compounds. One approach to defining the relationship of compounds to each other uses a chemical space defined by a set of 500 compounds described with 60 molecular descriptors onto which novel compounds can then be mapped.116 Another approach that has been described uses a rule-based classification scheme to define the common scaffold or core of compounds.117 This approach has been taken further to develop a scoring function to allow the prioritisation of compound libraries for testing.118... 

If the rank of M is less than M and N, then some elements of the matrix S will be zero, and the number of nonnegative elements will equal the rank of M. SVD has been used successfully to reduce the dimensionality of descriptor space of chemical libraries." Xie and co-workers report mapping chemicals with 10% of high-dimensional distances into two dimensions ranging from 30 to 100%, depending on the types of compounds and descriptors used. 

Many reports have been published addressing various aspects of diversity analysis in the context of chemical library design and database mining. Methods have also been published to map molecules from a high-dimensional chemical space to 2D/3D space so that direct visualization of chemical similarity and diversity becomes feasible.Fow-dimensional diversity space also... 

We established [54] that, in conjunction with 2D-based descriptors, ChemGPS can provide global chemical space coordinates by performing extensive comparisons with GRID-based principal properties for heteroaromatic compounds [59], principal properties ( z-scores ) of a-amino acids [60], as well as by comparison to several local PCA models. The initial ChemGPS map turned out to be 9-dimensional [54]. 

---