Key points are not available for this paper at this time.
A computer program that automatically classifies and clusters structural fragments extracted from mining the Cambridge Structural Database is described. The methodology is based on cluster analysis and multivariate data processing of distance matrix information describing the extracted fragments. Coupled with the calculations is a set of visualization tools that enable the user to view and verify the proposed classification scheme, and further explore it in varying levels of detail. Two examples are presented: the first is based on a simple difluoroalkene fragment and the second, more complex, on a chiral vicinal dialcohol, R 1 (OH)CHCH(OH) R 2 .
Barr et al. (Thu,) studied this question.