Metabolic Compound Database

Peter Karp pkarp at AI.SRI.COM
Tue Oct 1 23:34:03 EST 1991



Announcing

		  CompoundKB -- A Database of the Chemical Compounds
			   of Intermediary Metabolism


The CompoundKB database describes 981 chemical compounds of
intermediary metabolism.  For each compound the database lists:

	o Synonyms for the compound name
	o The systematic name of the compound
	o The CAS registry number
	o The chemical formula
	o The molecular weight
	o The chemical structure
	o Two-dimensional display coordinates for the structure

The CompoundKB is available in three different database formats to
facilitate access by end users: in a Knowledge Base form, as ASCII
delimited tables that can easily be loaded into a relational database
management system, and in the ASN.1 data-exchange standard format.
Although no special-purpose software is provided for searching the
CompoundKB, users can employ general software tools such as relational
database management systems and ASN.1 software utilities for
manipulating the database.

The CompoundKB is available online via the Internet from the data
repository at the National Center for Biotechnology Information.  Use
the Internet file transfer protocol (FTP) to connect to host
ncbi.nlm.nih.gov using username "anonymous" and your name as password.
Then connect to directory repository/metabolism/compounds/asn1 , and
retrieve all of the files found there.  This directory contains the
ASN.1 representation of the dataset; to obtain the RDBMS or
knowledge-base representation, substitute  rdbms  or  kb  for
asn1  in the preceding directory name.  The Lisp-language
software that the author used to develop the KB is in directory
lisp.

The database has been reviewed for correctness by a number of
scientists, and a manuscript (submitted to CABIOS) describing the
CompoundKB documents its error rate, the methodology by which it was
constructed, the semantics of the data, and the different database
formats.  The manuscript is available in subdirectory Doc of the
compounds directory on the NCBI repository (in postscript and latex
formats), and from the author.

Please direct questions or error reports to the author at the
following address:


	Peter D. Karp, PhD			pkarp at ai.sri.com
	Artificial Intelligence Center
	SRI International, EJ229		office: 415-859-6375
	333 Ravenswood Ave
	Menlo Park, CA  94025			fax:    415-859-3735




More information about the Biomatrx mailing list