ISP uses predictive computer models to investigate new ingredients

The New Jersey-based company has formed an alliance with Cellworks Group to develop new skin care approaches using the latter’s computer models.

Skin pigmentation was the topic of the pilot study using Cellworks Group’s Virtual Skin Platform, with the aim of modelling the effects of two ingredient projects on skin color.

Proteomics-based approach ​

The computer models are based on proteomics – the large scale study of proteins, their structures and functions. Modelling the interactions between proteins in the cell and how this would be affected by an external ingredient can help predict the action of the ingredient before launching laboratory tests.

Because it is cheaper and less time consuming to model these interactions in silica ​(in computer models), as opposed to in a laboratory, thousands of scenarios rather than just two or three can be investigated.

According to ISP, the pigmentation study assayed hundreds of bio-molecules, examined different application rates and allowed the company to predict effects before heading to the laboratory.

For example, the percentage concentration of an ingredient is one characteristic that can be predicted by these models, explained VP of research and development at ISP Dr Claude Dal Farra.

In addition, in the skin pigmentation studies the models can predict how different skin types will respond to the ingredients, he told CosmeticsDesign.

"It can also be very useful to tell you if you should expect side effects from the combination of various ingredients,"​ he added.

According to Dal Farra, ISP has been working with Cellworks to fine tune the predictive power of the models and is happy with the result.

"Yes many pathways are interacting, but we always check some of the predictions after the work is completed by Cellworks and we fine tune what has to be fine tuned,"​ he said.

Cheaper and more convenient​

In addition to being cheaper than using cultured cells in the laboratory, the system can also give freedom in terms of time.

“The co-culture equivalent would require primary cell lines that last 48 hours at best, imposing high costs and extreme time pressures,”​ he said.

For Cellworks Group the collaboration opens up a new market for the company as well as validating the predictive accuracy of the model.

“These ‘blinded’ validations are a very critical component in quantifying the accuracy of the predictive technology that has been developed,”​ said Pradeep Fernandes, Cellwoks Group’s president.

The pigmentation project was the first of three organized as part of the alliance between the two companies. The second will investigate anti-aging ingredients and the third will focus on synergies between different ingredients.