Can ‘drag and drop’ cancer simulation speed up drug discovery?

New ‘drag and drop’ computer modelling of key signalling pathways in cancer cells may dramatically reduce the need for ‘wet’ lab experiments to bring truly personalised treatment to patients.

Drug discovery scientists at AstraZeneca are starting to use a cloud-based simulation tool to bring alive the millions of potential changes in cell signalling that make cancer cells multiply uncontrollably.

On their computer screens, they ‘drag and drop’ cells, genes and proteins. Next, they draw in the basic signalling pathways that can go wrong in cancer cells. Then it’s over to computers capable of performing eye-watering calculations to fill in the gaps and predict what happens if different steps are blocked with drugs.

“We model the way many different proteins interact and we simulate experiments that we previously did in the lab.  We ‘test’ the likely effects of our drugs at different places on the pathway and identify the most promising places to intervene,” explains Jonathan Dry, Principal Scientist and global strategy lead, Bioinformatics, Oncology, Innovative Medicines and Early Development at AstraZeneca.

This ground-breaking advance in drug discovery methods is the result of a highly successful collaboration between AstraZeneca IMED and Microsoft. The original idea for the computer tool, called BioModel Analyzer (BMA), came from Jasmin Fisher, Senior Researcher at Microsoft Research and Associate Professor of Systems Biology at the University of Cambridge: “As a biologist myself, I wanted a user-friendly method of investigating the complexity of signalling pathways in a much more systematic way than was possible in the past. So we spent a lot of time and trouble converting heavy computer science into ‘one click’ analysis that would present information in a visual, graphical way for experimental biologists.”

It was worth the effort! BMA is one of several simulation tools being used to predict drug effects on cancer cell signalling pathways. But, says Dry, it works in a simple, logical way and is much easier to use routinely than some of the more ‘mathematically heavy’ systems being developed by other research groups and software companies.

“The Microsoft system captures the way that the biologist thinks about the pathway instead of trying to fully mathematically model it, and that suits the way we work,” he says.

Getting rid of the guesswork

Without computer modelling, scientists have to rely on an incomplete understanding of the stages in cell signalling pathways and their own intuition to choose which steps to block.  

“It’s like a huge jigsaw and, even if you have the pieces, you don’t know how they fit together,” says Fisher.

Dry explains that there may be 10s to 100s of different possible points to choose from in a single pathway. “Even with high throughput screening, that could take weeks of work but with a computer simulation you can do it in minutes,” he says.

Faced with a myriad of different choices, scientists currently tend to choose their favourites to investigate and may therefore miss better options:

“The beauty of computerised modelling is that you remove the current limitation on the number of hypotheses you can investigate and the potential for human bias,” says Dry.

Targeting leukaemia

The initial focus of the IMED-Microsoft collaboration has been on a type of blood cancer, acute myeloid leukaemia (AML). Dry explains that to understand the subtle differences in signalling between different AML cell lines, it is essential to get real data from real cells.

IMED scientists collected genomic, transcriptomic and proteomic data from AML cell lines in the lab. At the same time, Microsoft computer scientists developed highly sophisticated algorithms to model all the possible variations in a key signalling pathway and their likely outcomes.

This information was then used to simulate experiments to predict how drug combinations might be used to overcome resistance to drugs that target cell signalling in AML, and how this might vary between cell lines. The in silico data generated by the computer model were compared with those from in vitro tests in the laboratory.

Click image for a closer look

“Even I was surprised by the accuracy of the simulation!” laughs Dry. “It told us which drug combination would make a cell sensitive to treatment and the protein changes that led to that cell becoming sensitive.”

To ‘train’ such a model requires a huge data set on the signalling pathway of interest from multiple diverse cancer cells. As Dry points out, it is time consuming and not something that is done routinely. However, as the AstraZeneca/Microsoft collaboration continues, specific oncology projects, for example, in breast cancer, are being prioritised for the simulation approach.

Collaboration at its best

The closeness of the IMED and Microsoft teams, both based in Cambridge, has been a huge advantage to the BMA initiative. Dry explains that working with Microsoft is a true collaboration with a real exchange of knowledge, data and expertise:

“We contributed the big data sets and knowledge about the signalling pathways in order to fine tune the models, and Microsoft brought the software expertise. They captured the relationships between proteins in a pathway to build the model and run the simulation. We were then on hand to validate the results of simulations with our laboratory data.”

Fisher agrees about the value of such a united approach:

"We worked together every step of the way, everyone using their own expertise. As a result, we have real pre-clinical data that can impact the way drugs are developed. It’s genuinely meaningful, not just another piece of ‘blue sky’ research. It’s a super-exciting time!"

Towards greater precision

Modelling signalling pathways in all their variability, and simulating the effects of drug combinations at key points for individual patients, has the potential to move personalised medicine well beyond its current boundaries.

Current treatments target single points in signalling pathways, based on genetic mutations in a patient’s tumour. But future personalised therapy is likely to focus on multiple points in the pathway initially identified by computer simulations and confirmed in the laboratory. Ultimately, such simulations could even take account of interactions between the tumour and its broader environment, including host immunity and DNA damage response mechanisms.

“We’re moving to a whole new level of understanding of the intricacies of drug targeting,” concludes Dry. “We’re moving away from hitting a pathway just because we know it’s active. Instead, we’re identifying the best place to hit the pathway in order to get the best possible response for that patient. For AstraZeneca, there is also the potential to improve our drug development success rate – by focusing our oncology research on the targets that matter most.”

Page Atlas ID: 1008544.011
Date of next review: September 2017