Biodiversity Informatics

Natural Substances in the Compositae: The Bohlmann Files

Challenges and Solutions

For practical reasons the chemical editing (carried out by a sub-contractor) and the nomenclatural editing and database development (at the BGBM) were done at different places. Therefore, the chemical data entry and updating was continued in the ISIS database and the parsing process had to be repeated before finally publishing the completed data. The Bohlmann Files is currently a centralized database. Distributed editing of different components has not been within the scope of the project, but will be needed to update and correct the data in the future. Implementing a WWW-based remote editing facilities is possible and will have to be a priority for a follow-up project.

With respect to botanical names, the assignment of the status (accepted or synonym) may often be a matter of scientific opinion and it may be changing with the progress of research. The Bohlmann system implements a internally consistent taxonomy using correct names (in the sense of the code of botanical nomenclature), and it tries to accommodate commonly used synonyms. Where necessary, a pragmatic decision was taken, e.g. when current entries in relevant databases or monographs provide contradictory information. However, the system can incorporate new classifications at any time, without loosing the separately kept original designation (table tdChemistry.Vorkommen) stemming from the chemical literature.

The commercial software product Accord caused a major difficulty on the technical side. It had been used for the incorporation and querying of chemical structures and it was - at the time of the start of the project - the only reasonably priced solution for license-free web publishing and particularly on-line querying of chemical structures (Accord WebDriver). The latter aspect is an essential part of the Bohlmann-Files project. However, during the project period the company developing this tool went out of business. While parts of their projects were taken over by another firm, the WebDriver development was stalled. As a consequence, updates were not following the development of the rest of the Accord product series. The considerable difficulties this caused in the setting up of the web interface finally resulted in switching to a different software package, JChem from ChemAxon. This process included a complete re-implementation of the web-pages and a migration from Access 2000 to MS SQLServer - a step that had not originally been foreseen for this phase of development. However, this resulted in far better possibilities for formulating search requests (substructures, similarity search) and includes the potential for remote editing (editing, inserting and deleting of chemical structures via the internet).