High-order pattern recognition

There is a tendency to think of medicinal chemistry as primarily a logical exercise. A specific and trivial example would be the much maligned QSAR exploration of methyl, ethyl, butyl, futile. This author believes that equating medicinal chemistry with QSAR is incorrect. There is a definite place for what might for want of a better term be called high-order pattern recognition. A specific example is the time tested 

---

The hypothesis of a normal distribution is a strong limitation that should be always kept in mind when PCA is used. In electronic nose experiments, samples are usually extracted from more than one class, and it is not always that the totality of measurements results in a normally distributed data set. Nonetheless, PCA is frequently used to analyze electronic nose data. Due to the high correlation normally shown by electronic nose sensors, PCA allows a visual display of electronic nose data in either 2D or 3D plots. Higher order methods were proposed and studied to solve pattern recognition problems in other application fields. It is worth mentioning here the Independent Component Analysis (ICA) that has been applied successfully in image and sound analysis problems [18]. Recently ICA was also applied to process electronic nose data results as a powerful pre-processor of data [19]. 

For all those above reasons, trying to classify environmental odour sources in real life with pattern recognition techniques gives rise to a spread of observation points in the different clusters. For the signal processing, that implies to have a great number of samples in order to consider the various conditions and chemical compositions. But that induces a high dispersion of the data of a same class and/or of the same concentration. 

ADAPT has a variety of pattern recognition and statistical methods available for use. The object of the analysis phase is to find discriminants that separate subsets of the data into the proper categories. That is, one is trying to find mathematical models that will classify compounds as belonging to the active or inactive subset based on the molecular structure descriptors available. This phase of SAR studies Is guided by the user in a highly interactive manner in order to search through the available descriptors for the best set. 

The highly influential Rule-of-5 paper by Lipinski et al. was developed in order to provide medicinal chemists with calculable parameters... that were likely related to absorption or permeability, presented in a way that would focus on the chemists very strong pattern recognition and chemical structure recognition skills . The authors analyzed a set of 2245 compounds from Phase II and later, using parameters previously indicated to be related to these properties MW, related to intestinal and blood brain barrier... 

The overriding conclusion from the extensive data available in the CSD is clear. In order to convince two different discrete chemical species to coexist in a molecular co-crystal there needs to be some specific molecular-recognition based reason for their solid-state union [93]. Although individual structures that defy rationalization will appear from time to time, there is no doubt that the important big picture reveals structural trends, patterns of behavior, and reproducible motifs that, when combined, can be developed into a library of high-yielding supramolecular reactions. 

---