Beneficial bug: AI predicts how molecules will react with skin microbes to ensure product safety

The skin is our largest organ and the most exposed not just too topical personal care products but also environmental pollutants.

Dr. Vineet K. Sharma, Associate Professor of the department of Biological Sciences, Indian Institute of Science Education and Research Bophal, explained that the skin also harbours microbes can affect how the skin reacts to the pollutants.

“Our skin is the most exposed human organ. When we are exposed to pollutants like cosmetics, there is a possibility that these microbes can reduce the efficacy of a product or covert it into a toxic or carcinogen. Environmental pollutants can also be converted on the skin and become more toxic,” ​said Sharma.

SkinBug allows researchers and developers to predict how a molecule will react to the skin microbes and determine the likelihood that it will be metabolised and bio-transformed it into something else.

The team, which primarily researches on the skin microbiome, developed this AI software to address what they believe is a significant problem for skin and health.

“This software is fed with more than one million reactions and it can predict how a molecule can be metabolised by the skin. It can predict if this molecule can be converted into a toxic product or a carcinogen. It’s important from a health and skin care perspective,”​ said Sharma.

He added that an AI tool like SkinBug was indispensable in determining how molecules will react to the skin, which he stressed was a very difficult task.

“Scientifically, it is a very challenging and important question and we thought AI could help to handle this question. Because this cannot be handled by a simple approach or experimentation for each molecule. It's time-consuming and tedious.”​

Sharma believes that SkinBug holds a lot of commercial potential. While he declined to name the companies, Sharma said the team will be providing two ‘topmost’​ companies with a version of SkinBug with additional features.

“This will be something they require so they can quickly do what is required for their own developments.” ​

A personalisation tool?​

As more information on the skin microbiome comes to light, the team hopes to expand the use of SkinBug.

“Our knowledge about skin microbiome is continuously increasing. As we get more and more information from experimentation and worldwide studies, I think one day we can envision – and this is something we are also working towards – that it can be used for personalisation,” ​said Sharma.

He elaborated: “You could be able to do your own microbiome test at a very low price. This software could be an app or online. You can use it to know more about your skin type, what products would suit you and how different you are in terms of skin microbiome.”​

 ​

SkinBug: an artificial intelligence approach to predict human skin microbiome-mediated metabolism of biotics and xenobiotics​

Source: Pubmed​

Authors: Shubham K. Jaiswal, Shitij Manojkumar Agarwal, Parikshit Thodum, and Vineet K. Sharma

Doi: 10.1016/j.isci.2020.101925