Bibliography
Chemical Information Retrieval and 3D Searching

The bibliography provided here is a growing concern, compiled from past and current readings in the following areas:

Bibliography files

The bibliography is too large to place in one single text file -- it has over 2650 individual citation entries -- so has been broken up into 9 separate segments. Each segment is organized alphabetically by the leading authors last name. Beware of unusual diacritical mark translation (see explanation in section below).

Why these references were compiled

Some of the citations herein have been incorporated into a review article entitled "Chemical Structure Handling by Computer", now published in the Annual Review of Information Science and Technology (v.32, pp.271-337, 1997).

Bibliography database format

The material listed here is a translation from the contents of a Macintosh EndNotes v2.x reference manager database containing over 2650 citations for journal articles, book chapters, abstracts, patents, books, etc. Several of these references point to active online WWW sources where the actual text may be browsed (do a FIND for "awod", for example, in any of these subsidiary bibliographies and find all the citations with online references to the electronic journal NetSci). Each reference below has a specific "entry point" and URL address that will persist even through multiple versions of this bibliography, and therefore can be cited (as needed) across the Novartian Intranet web.

Diacritical marks and spelling

Herewith, an explanatory note about diacritical marks and spelling in non-English names. It identifies some printing and visualization problems with diacritical marks that do not survive the translation from complex PC/Mac character sets into the more simplistic world of HTML. This is obvious, for example, when one examines the entries for Guner:1986 and following (normally spelled with u-umlaut, and present in this text as part of the unsupported extended character set (decimal 237) and often visualized by web browsers as a capital-Y) or Carbo et al:1990 and following (normally spelled with an o-acute, and often visualized by web browsers as a hyphen). In these cases, the diacritical mark does not translate properly into HTML, concealing the proper spelling of the author's name.

Most vowel diacriticals, and a few consonant diacriticals, as would be found in common German, French or Spanish are supported in the MSWord document in their native format (e.g.: Guner, Ariens, Laszlo, Bohm, Zoltan, Carbo, Veronique, Rose, etc.), but, as seen above, are NOT supported in the default HTML formats. Russian and Ukranian names are transliterated in canonical form, following the citation practice of ACS and CAS (as found in the CAS files on STN and DIALOG). However, for Hungarian, Czech and several other Eastern European name forms, the typefaces supported by MSWord and standard Windows/Mac font suppliers seldom allow the construction of such proper forms as "hacheck" (c with inverted-^) or acute-c. In the former case, such as with authors Novic(inv-^) and Kvasnic(inv-^)ka, the diacritical is explained in parenthesis following the consonant. The the latter cases, such as with authors Randic and Trinajstic, the diacritical marks are standalone and precede the consonant. With apologies, many other diacritical marks are simply ignored. Chinese, Japanese and Korean name forms have been re- ordered, as necessary, to follow the American and European convention of placing the family name in primary sort position.

Spelling and diacritical mark variants, when discovered to be different in different database records, have been resolved by reference to the original papers.

For primary sort order, fused diacritical marks have been ignored.

greg.paris@pharma.novartis.com
cgp@res.nibr.com
cgp@scils.rutgers.edu
Converted: 19-May-97
Most recent update: 14-May-99