Solo disponible en BuenasTareas
  • Páginas : 14 (3285 palabras )
  • Descarga(s) : 0
  • Publicado : 29 de agosto de 2012
Leer documento completo
Vista previa del texto
About the concept of Chemical Space: a concerned reflection on some trends of modern scientific thought within theoretical chemical lore1.
Ramon Carbó-Dorca2

Sir! In a recent paper [1], a footnote was written by the present author as coauthor in order to comment about the concept of Chemical Space, which is a term appearing sporadically since old times in some available chemical literature.The mentioned footnote text was printed as follows, with the underlined part used to emphasize here some relevant parts:
Often in the QSPR or QSAR common literature the term chemical space appears. This concept it is not currently used with the same sense as a possible molecular space notion or the equivalent term: space of molecules given here, but in a generic fuzzy manner, sometimes even usingmeaningless and preposterous literary ways. The authors would like to find within the QSPR literature a sharp definition of such chemical space term, but until the present time found none. So, in order to employ the terminology of the present work as far away as possible from the usual classical QSAR fuzzy one employed by some authors, the term “space of molecules” has been chosen and will be usedhere, although “molecular space” might be also synonymously used. This notion has been and will be also employed here as a companion of the term: space of parameters or the synonymous concept: space of descriptors, which is also a welldefined definition with a precise meaning, commonly used in the task to build up a set of discrete linearly independent vectors, empirically describing a MPC. Fromsuch a vectorial descriptor space construct it is normally obtained a QSPR linear equation intently employed for molecular unknown property estimation purposes. The statistical techniques, employed upon the vectors of the space of descriptors, within the empirical QSPR equation search algorithms, which drastically reduce the dimension of molecular descriptors’ space, generate the dimensionalityparadox.

The writer of this letter subscribes the ideas of the footnote as they are. Definitions about what Chemical Space must be thought nowadays are in

This letter is written to the memory of Professor Odd Gropen, a faithful friend, a dedicated scientist and an occasional watcher of Star Trek. 2 Institut de Química Computacional Universitat de Girona Girona 17071 Catalonia (Spain)e-mail: Phone: +34-972-418357.

essence within this small piece of work. But the present author thinks it is worthwhile to surf into some sources of information, which will lead to a fuzzy intellectual panorama. The subscriber of this letter is eager to say that it is necessary to describe such a way of thinking, as a warning in a travelling research track, which mightlead to poor science background building in 21st century. When talking about the difficulty of finding a sharp definition of Chemical Space, the authors of the previous footnote wanted to express the fact that no definition appearing in the literature seemed to be satisfactory. The basis of this opinion is tried to be explained here. However, an even more exhaustive analysis of the inquiry, in orderto clarify the questions awakened in the quoted footnote, has aroused in the present author a deep concern about some contemporary scientific opinion trends, hopefully thinking they are just shared by few. The present letter is a consequence of this state of mind. On the other hand, and to start this journey in the Chemical Space quest I have found that in the Wikipedia the search engine presentedsome site with the following text:
Chemical space is the space spanned by all possible (i.e. energetically stable) molecules and chemical compounds – that is, all stoichiometric combinations of electrons and atomic nuclei, in all possible topology isomers. Chemical reactions allow us to move in chemical space. The mapping between chemical space and molecular properties is often not unique,...
tracking img