Skip to content

Terms and Conditions for utilizing LEO Pharma Open Innovation web service for AI-prediction of molecules’ biological activity based on chemical structure

Ultra-brief caption of the ‘terms and conditions’

LEO Open Innovation’s Artificial Intelligence (AI)-powered prediction model requires the submission of the molecule’s structure which will be treated as secure and confidential data as best as we can. LEO Pharma will not claim, disclose, store or manipulate the data other than predict and return the potential biological effects, unless permitted by the user to do otherwisw. The terms and conditions are as follow.

The provided service

Based on a molecule’s chemical structure, the service provides a prediction on how likely such a molecules would have a biological effect. On a secure website for public access, the user submits a molecule’s structure to a prediction software that returns the results over the internet. The prediction software utilized Artificial Intelligence (AI) or Machine learning (ML) that has been trained using real laboratory results from thousands of molecules in biological models provide by LEO Pharma Open Innovation.


Our intention is to provide an easy way to engage with the science we explore trying to help people living with skin conditions. The service aims to reduce the complexity and distance between a theoretical molecules and laboratory results, making it easier to explore many different molecule’s ability to reduce skin inflammation.

What you get

Free of charge access to an AI model that can predict molecules’ ability to inhibit inflammation relevant to skin diseases, which includes the condensed results from LEO Pharma’s testing of thousands of molecules in biological models.

How the results could be used

The predictions can be used to e.g. aid the design and synthesis of new molecules, both by providing data on potential cytotoxicity (TOX) to be avoided, and pursuing or confirming biological activity of novel structures. Furthermore, the predictions can be used to filter and select a few molecules from a list of many, that could be submitted to real testing in the LEO Pharma Open Innovation laboratory.

What if the molecules are predicted to work?

A molecule that is predicted to work and has a low degree of cytotoxicity could be relevant to test in the real biology laboratory. However, most molecules are very far away from being relevant drug candidates and a multitude of additional laboratory testing is required to evaluate further potential. Read more on the website on what could be done scientifically to follow up.
Also, molecules which activity can't be predicted could be even more valuable to test and potentially feddback into the AI-prediction software. However, from a practical perspective it is not always possible to acquire or synthesis a completelty novel molecule.


The AI predictions are always only predictions regardless of the degree of confidence and may not actually correlate with the results from an actual test in the laboratory.


The service is provide free of charge and as-is and may be subject to change and availability completely to LEO Pharma’s discretion. If the prediction software is under heavy use there might be long delays in retrieving results.

Our claims

LEO Pharma makes no claims on any data submitted or returned as predictions.

The risk involved

We aim to provide a secure process as possible while transferring the molecular structure from the web interface to the prediction server, but we have limited control of the infrastructure or services provided by others that provides access to the internet and can therefore not guarantee that the submitted and predicted data remains confidential. There is no deliberate disclosure step in this process and the molecules structure is not published as in being disclosed to the public. Also, the data is submitted using hypertext transfer protocol secure (HTTPS) which provides a secure and encrypted internet communication.


The submitted molecular structure is only used to predict biological affect and the results is only returned to the user. The submitted structures or prediction results are not further evaluated, stored or communicated by LEO Pharma or anyone else. We will do our best to keep your data confidential, but we cannot provide any absolute guarantee of confidentiality.
The confidentiality obligation does not extend to information that is already in the public domain, or becomes part of the public domain by disclosure through no violation of this agreement, or if LEO Pharma has been lawfully in possession of the information, or is independent developed by LEO Pharma or any affiliates.


As there is not public disclosure the utilization of this service should not affect any possibilities to patent a molecule. Although, as mentioned under “risks” there is no absolute control of other parties’ access to the information, as with any usage of the internet.

We sincerely hope you will explore the vast options in chemistry to identify new ideas for molecules that can ultimately be used to treat inflammatory skin diseases.

Scroll to top