|
|
||||
| Home | Services | CDISC | News | About us |
Web-enabling & web-services |
i18n and XML |
Molecular Modeling |
SOAP enabling of a number of Bioinformatics packages at a research institute, using Perl, C and C++, as well at client side (service consumer) as at server side (service provider)
Development and
implementation (in Java) of a neural
network for the prediction of biodegradability of surfactants. The
neural network was implemented as a Java applet and made available at a
password protected web site through the corporate's intranet.
Development of a Java applet implementing
the theory of Miller and Macosko for the prediction of properties of
polymeric networks (thermosets), as in coatings (paints) and adhesives.
This tool was made available on the intranet of the worlds largest
coatings company.
Also for an alternative theory (Bauer), a Java applet was constructed and
deployed on this company's intranet. Specific for the Bauer-theory is
that it takes into account the molecular weight distribution of one or
more of the components (be it polymer or crosslinker), as is the case in
many practical systems.
Development and web-deployment of
software for the prediction of Hoy's
δ-parameter for chemical substances. This software increased
the efficiency of research at a major coatings company considerably.
Before the web-enablement of this software, researchers at this company
performed the calculations using a spreadsheet program, which was very
user-unfriendly (leading to many errors), and of which different versions
existed (uncontrolled software deployment).
Analogously, an web application was developed for the prediction of
compatibility of possible plasticizers with PVC. Before the availability
of this application, researchers at this chemical company in most cases
had no idea whether their envisages plasticizer would mix with PVC or
not. It is clear that this application increased the efficiency of their
research considerably.
Both these applications were soon regarded as "killer applications" within the company.
Development
of a web-interface for software for the prediction of hydrophobicity and partition
coefficients (XLogP). The software was only available on expensive
SGI workstations (so that only a few people could perform the
calculations). After the web-interfacing all researchers within the
company could simply sketch a molecule on their PC, and then with one
click perform a partition coefficient calculation.
Development of a web interface for MOPAC, the well known semi-empirical quantumchemical software. The calculation itself is performed on the server, and when it is ready, the submitter automatically gets an e-mail with the results. At the same time, an e-mail with details of the calculation is send to the company's quantum chemistry specialist for inspection.
Development and web-enabling of a neural network for the prediction of CMC's of surfactants.
Web-enabling of a QSAR for the prediction of toxicity of a peticular class of chemical compounds. This QSAR was developed by a coworker of the toxicity lab. After this coworker left the company, noone could use the QSAR anymore as it was embedded in a special package which was mastered only by this coworker. We then translated the QSAR in a Java program and then made the QSAR available through the intranet, so that the users only needed a browser.
Development of a web-tool for the calculation of vapor pressures using a data base with Antoine-coefficients.
Active member of several workgroups (ODM development, Compliance, Educational) at CDISC, the organization for the development of new standards for Clinical Data Exchange.
Introduction of XML and Java to the central Information Services Department and Information Management of a pharmaceutical company.
Co-initiation of first XML-submission to EMEA of a complicated electronic dossier.
Co-initiation of a project for the development of a web-application for translation of texts for pharmaceutical information leaflets, in many (also non-Western) languages and character sets, and stored into a database in XML-format (internationalization).
Within Akzo Nobel, many research problems were solved with the help of molecular modeling, in many cases leading to new or improved products (coatings, fibers, chemicals).
Development of a new method for the prediction of enantioselectivity of chiral selectors.
Development of new methods for the prediction of possible crystal structures of polymers. The existing molecular modeling packages did not provide this functionality, so we developed it ourselve.
Development of a number of methods for predicting network properties of thermosets.
