A quest in the cloud

The Story

When the Cerebral Palsy Alliance (CPA) approached us in late 2020, they wanted something revolutionary: a data platform that would unlock the future of cerebral palsy research for scientists around the world built around digging deep into the genome. The challenge was daunting, nothing like this had ever been accomplished before –and there had been several false starts. Our team jumped into the project head first.

The opportunity to do good at scale was clear. Gene research had emerged as a critical path to curing Cerebral Palsy. Because of this, if global service providers and researchers could share their findings in one centralised location, progress against cerebral palsy would be super-charged. And future research would also be super-charged, because the data would be available for the application of new Artificial Intelligence and Machine Learning analysis.

To deliver on this promise, CPA knew it had to go well beyond simply facilitating access to global genomic datasets on this disease. Our team was tasked to achieve several additional outcomes:

Enable researchers and medical professionals to more easily gather insights and data, as well as share their own findings.
Speed up the discovery process.
Promote fast communication and global collaboration on genetic findings.

The Challenge

First, let’s take a step back. All of us have likely heard of cerebral palsy. It would be easy to gloss over what CPA and Kablamo are fighting –so here it is. Cerebral palsy is the most common motor disability in childhood. It devastates individuals and families, and causes a long list of life-long challenges, including the need for special care.

Tasked with developing a new ambitious tool in this fight, we utilised our design practice to understand the key product users (KPU) and develop a platform that would best cater to their unique needs. Human centred design informed every step since experience has taught us that the most powerful tool is useless if it isn’t used, or used to only a limited degree or by the wrong people. While we initially considered clinicians as potential users, the project’s time and financial constraints meant it was more practical to isolate the platform to scientists and researchers, with the opportunity to broaden scope for other users at a later stage.

Once we understood our KPUs, we needed to develop the platform around solving their biggest hurdles: cost and time. Being able to draw on data collected by their colleagues is essential for improving the efficacy of cerebral palsy research. But with each ‘family’ of data costing $10,000 and taking upwards of a year to collect, these financial and time constraints could hamper the progress of any breakthrough.

Innovation

Centralised data portal with increased functionality

Integrity

Fostering a culture of global collaboration and sharing

Intelligence

In-depth understanding of key product users (KPUs)

The Solution and The Result

With the CPA having already experienced multiple unsuccessful attempts at building this platform, it was important that we approached its design with their expectations and user needs front-and-centre. Building trust was critical in our journey to this world-first medical research platform. We conducted a series of workshops to establish requirements, brainstorm as a team and ensure that all parties were aligned. We created multiple personas to better clarify the platform’s intended audience, as well as to outline all the various user problems that would be addressed in the solution’s initial release.

Given the time-sensitive nature of the project, we knew it was critical to prioritise the essential features and minimise the non-essentials. Ultimately, this allowed our team to concentrate on perfecting the platform’s core functionality without compromising the user-friendly experience and having a practical, intuitive interface.

Prior to engaging us, the CPA had been waiting 18 months for a solution. We delivered the successful product – called CP Commons – in just eight weeks. What was previously considered impossible is now a functional platform that allows the sharing of critical cerebral palsy research between researchers and scientists from around the world.

One of the solution’s core components is a centralised data portal. A significant advancement is its ability to manage large files, which enables users to search for and request explicit data, create projects and upload their own data. We also delivered an administrative portal that enables the client to regulate user access to the primary database platform – a vital component given that researchers must be screened and approved for access due to the confidential nature of the information.

CP Commons is now helping to foster a culture of global collaboration and sharing in the cerebral palsy research community.

Technologies

AWS: We chose AWS for its scalability and affordability. S3: An affordable and robust service, S3 is used as both primary storage for the uploaded genetic data, as well as for hosting and serving the front-end application. Data is first placed in a landing bucket with a set expiration policy before being copied and stored permanently in the data bucket. The files, which can reach several gigabytes, are managed using direct S3 file uploads with V4 signed requests. RDS: RDS is used for storing metadata describing genetic data stored in S3, as well as for managing users, projects and groups. Fargate: We selected Fargate to run the API and manage all the background copy tasks for transferring data from the landing to data buckets. SQS queues are used to handle these processes, including metadata validation and storage in RDS, followed by asynchronous processing independent of user activity. The API itself was written to be stateless.

This project will accelerate the discovery of how genes contribute to cerebral palsy and put a valuable new tool in the hands of the world's best and brightest CP researchers.
- Rob White
Cerebral Palsy Alliance CEO