ACTION on cancer

Lead Research Organisation: University of Manchester
Department Name: Computer Science


Death from cancer is typically both slow and painful, and few families have been spared its scourge. Cancer is also one of the world's greatest killers (13 million deaths and 22 million new cases per year by 2025), and it is estimated that every second person on the planet will develop cancer at some stage of their life.

Over the last 30 years our knowledge about cancer has increased enormously, and now, for the first time, we understand the fundamental nature of the disease(s): malfunctioning in the way that cancer cells process information. All the cells in our bodies process information about their internal state, and communicate with their neighbours, and when this goes wrong cancer can occur.

As everyone's cells are different, and there are very many different ways that this information processing can go wrong and cause cancer, it is not possible to design a single treatment for cancer, or even for a sub-type of cancer such as breast cancer. Instead what is needed are personalised treatments tailored to each patient's cancer. However, such personalised treatments are very expensive to design, and the expertise to do so is limited. In addition, it is often necessary to execute custom designed experiments to better understand what is the best treatment. Therefore the only way to make personalised cancer treatment available to everyone is through laboratoryautomation, and the use of artificial intelligence (AI).

In this project we will develop ACTION, which will be a prototype AI system for the design of personalised cancer treatments. ACTION will focus on chemotherapies - design of drug cocktails. Given initial information about a cancer ACTION will extract all the relevant knowledge it can find about the cancer, both from databases and computational models of cellular information processing that scientists have developed. ACTION will rationally integrate this knowledge, and infer what extra knowledge is required to make the best decision on how to treat the cancer. ACTION will then automatically execute custom designed experiments using laboratory robotics to determine the missing information. Finally, using all the knowledge it has gathered, ACTION will decide on the best chemotherapy.

We will evaluate ACTION using different types of cancer cells grown in the laboratory. This avoids the ethical complexities of working with patients, and is much cheaper and faster. If the development of ACTION is successful it will then move to testing with patient derived cancers.

Planned Impact

Impact Summary
The proposal is an ambitious one with a high potential for significant technological, Medical, economical, and societal impact.

Cancer is one of the world's greatest killers (13 million deaths and 22 million new cases per year by 2025), and it is estimated that every second person on the planet will develop cancer at some stage of their life. The annual treatment of cancer costs the world ~$900 Billion per annum (more than any other disease), and there is deep concern that cancer treatment is becoming unsustainable, as new therapies can cost > £100,000 per patient per year.

The Funding Landscape
The proposal aims take the concept of AI designed personalised drug treatment from TR2 (Principles demonstrated through experimentation) to TR3 (Early proof of concept demonstrated in the lab). The main objective measure of success of TR3 will be that ACTION will make automated recommendation on personalised cancer treatment that is as good or better that current 'best practice'

RDK has significant commercial experience. He was a founder of the spin-out company PharmaDM, which developed data mining tools for the pharmaceutical industry. RDK also consults for a wide range of multi-national companies. LS has less commercialisation experience, and she will undertake appropriate business and commercial training.

The applicants understand and have experience with the variety of follow on governmental and medical charity. For example, RDK was involved with Cambridge in a MRC Biomedical Catalyst project to develop an anti-malaria compound.

One year before the end of the project we will institute a meeting of the project Advisory Board to decide on an appropriate commercialisation pathway. To take ACTION to TR4 we will first seek proof-of-principle funding for this through the University of Manchester EPSRC Impact Acceleration Account. However, this will be insufficient to cover the high costs. We will therefore seek following on funding from Cancer Research UK, the MRC Developmental Pathway Funding Scheme (DPFS), etc. We will also approach our VC contacts.

The evidential support to produce a high quality follow on funding application will come from two main sources: publications, and Intellectual Property

The applicants have an excellent publication track-record, our papers have been published in Science, Nature, PNAS, etc. We will also attend and submit papers to relevant conferences.

Stakeholder Engagement
Through participation in Big Mechanism we are aware of the complex landscape of stakeholders in cancer. There is a growing enthusiasm for the potential of AI to help in cancer treatment and diagnosis, however this can only succeed if all stakeholders are included and involved:

RDK is engaged in cancer research in Manchester through membership of Manchester Cancer Research Centre (MCRC). To ensure a close relationship with MCRC we will invite Professor Andrew Hughes onto the ACTION advisory board.

To advise ACTION in industrial involvement we will invite Professor Kevin White onto the ACTION advisory board. We have worked closely with him in Big Mechanism. He is a US National Cancer Institute board member, and President of Tempus, where he oversees its scientific operation.

Patient involvement is also essential. Through Big Mechanism we have worked closely with Cancer commons: a 'patient-centric not-for-profit network of patients, physicians, and scientists that help identify the best options for treating an individual's cancer'. Dr. Marty Tenenbaum, Chairman of Cancer Commons, is excited by our research, and we will ask him to be on the Advisory Board. He is both a renowned computer scientist, and a cancer survivor.

Research capacity building
The project will train three PDRAs in key areas of future science and technology: AI, machine learning, bioinformatics, medical informatics, and cancer biology.


10 25 50

Related Projects

Project Reference Relationship Related To Start End Award Value
EP/R022925/1 01/04/2018 30/04/2020 £901,730
EP/R022925/2 Transfer EP/R022925/1 01/05/2020 31/03/2022 £616,470
Description We have developed machine learning methods to help decide on cancer drug therapy.
Exploitation Route Cancer treatmnet
Sectors Pharmaceuticals and Medical Biotechnology

Description We are working with Imagen Therapeutics on cancer diagnosis and treatment.
First Year Of Impact 2021
Sector Pharmaceuticals and Medical Biotechnology
Impact Types Economic

Title Closed-loop AI experimentation 
Description Our work is now causing a revolution in materials science. 
Type Of Material Improvements to research infrastructure 
Year Produced 2009 
Provided To Others? Yes  
Impact Closed-loop AI experimentation