M. Sonderegger et al. / Bioinformatics approach for data processing
83
have the ability to submit compound information from any networked computer equipped with a web
browser. No paper forms are needed.
3. Details
The web-based software can be broken down into five distinct sections: (1) information submission,
(2) mass analysis, (3) data review and (4) automated email notification upon analysis completion, coupled
with instructions to assist researchers when downloading their analysis results from the website. (5) An
additional section includes administrative tools such as database maintenance, statistics and accounting.
For security and accounting purposes every user is required to log onto the system. They then choose
the general category for analysis: proteomics or small molecules. New users are required to submit in-
formation such as email, name and phone number; for future reference, this information is stored in a
database. The scientist then submits all necessary information pertaining to the compound in a web form
(Fig. 2). For small molecules this information includes molecular formula and mass, solution concen-
tration and solvent, as well as compound purity and toxicity. In addition scan range, compound type,
functional groups and general comments can be submitted to assist the analyst when running their com-
pounds. For proteomics the web form requires the submission of the molecular weight and protein quan-
tity, purity and concentration. Biological source, sequence, possible modifications, storage instructions
and comments are also stored in the database.
Analysts in the mass spectrometry laboratory can also access the stored information from a web
browser where compounds are listed according to the analytical technique requested (Fig. 3). The cover
page of an automatically prepared PDF file contains all the information about the sample as well as
space to report analysis results (Fig. 4). Some embedded Java-scripts facilitate the automatic calculation
of the expected masses and the error on the observed mass. Mass spectral data are printed to a PDF file
and attached to the cover page. If necessary for further analysis, or when requested by a researcher, a
compound can be forwarded to another method of analysis.
After a final data review – which can be performed from any web browser – the data is stored on the
server for retrieval and the scientist is notified by email about the completion of the analysis. The file
containing all the compound information as well as mass spectral data is only about 100 kB in size.
Typically our system allows online access to the data for three months before it is permanently stored on
a CD.
In addition to improving efficiency, the paperless system has implemented some other tools to make
it easy for the user, those performing mass analyses and those interested in throughput. As all the in-
formation is stored at one central location, generating statistics about the performed tasks is easily ac-
complished. Detailed lists to track each laboratory as well as each user’s sample load can be generated.
A feature to send email messages to the users for the paperless system has also been implemented to
address just researchers interested in a single method of analysis or to reach everyone who submits com-
pounds to the facility. There has also been care taken to provide an online administration interface to
add or update user information and to change incorrectly submitted compound details. Submissions may
be deleted from the database completely or just removed temporarily if a researcher decides to send the
compound at a later time.
In order to make a system such as JULIAN functional, there are a few necessary tools. The basis is a
Microsoft NT4 Server installed on a PC system (Intel PIII 550 MHz, 256 MB RAM, 2 mirrored 13 GB
hard drives) which provides the Microsoft Internet Information Server (MS IIS 4). User and compound