Intelligensys Formulating Knowledge - neural computing in product formulation

Developing better products faster

Novel Applications

Our INForm and FormRules software packages have been used in some interesting ways for application areas outside mainstream formulation - here is a flavour of what can be achieved.

Modelling Complex and Indirect Relationships to make 'Long Range' Predictions

One big advantages of neural computing is the ability to model and hence predict non-linear cause and effect relationships between parameters which intuitively you might think are related but which are difficult to model. For example, the connection between the way a drug is formulated, and the way it behaves within the body. Or the relationship between baby teeth and adult teeth. Or understanding how light and nutrient concentrations affect plant propagation.

Models linking flask to plant

Some proven examples are:

Marcel de Matas and his group at the University of Bradford, UK, are predicting in vitro - in vivo relationships for drug inhalation. Relevant papers are given here.

Qun Shao at Bradford, working with collaborators from Beijing, has studied how stress in rats can be monitored by using biomarkers in their urine. A copy of the poster is available here. The Bradford group have also worked with Indian collaborators to look at genotoxicity for petrol pump workers - the full reference is here.

Professor Dang Van Giap in Vietnam, with two PhD students in dentistry, is using FormRules to investigate the relationships between baby teeth and adult ones (considering their shape and size parameters) and preliminary investigations look promising. They also plan to use INForm to predict the shape and size properties of adult teeth based on the data for baby teeth. This might be helpful for prevention of some dental problems, for example malposition.

Professor Pedro Gallego and his collaborator Dr Mariana Landin, working in Spain at the University of Vigo and University of Santiago di Compostela respectively, have looked at the in vitro culture proliferation assay with microshoots of kiwi fruit, examining the effect of light conditions and the sucrose concentration in the culture medium. The references for their publications are given here.

Reducing Costs and Making Better Use of Resources by Optimizing Processes

A key driver today is reducing both costs and, perhaps more importantly, the carbon footprint of various processes. An article in The Pharmaceutical Journal in December 2009 (page 623) highlights the growing importance of reducing the amount of energy consumed by pharmaceutical processes, and of course the same holds for processes in general.

Optimizing processes has been undertaken using our software in a number of areas, including the following:

AstraZeneca scientists from Macclesfield, UK have looked at finding the cause and effect relationships in formulation process optimization data for an immediate release tablet, finding which process conditions and intermediate properties are related to dissolution rate. Interactions between some of the parameters – which were difficult to find by a statistical study – were discovered automatically by our FormRules program. Find out more about AstraZeneca's use of our software here.

Jim McCabe from AstraZeneca has used INForm and FormRules to determine the process conditions that determine which polymorph is produced - a key issue for pharmaceutical scientists. The full reference is on our publications page. This sort of work is also being undertaken in the group of Professor Roger Davey at the University of Manchester, UK.