How tech is turbo charging healthcare research

By Amazon Web Services

From speeding up genomic analysis to improving drug development processes, AWS shares how AI and the cloud are accelerating researchers’ work.

Our DNA can tell doctors a lot about how to treat us better. Healthcare researchers are increasingly turning to genomic analysis to find out an individual’s risk of contracting a disease and detect conditions early.

But this knowledge is tucked away in our genes. Researchers need immense processing speeds and powerful algorithms to unpack the information, so doctors can deliver more accurate diagnoses and more precise care.

Amazon Web Services (AWS) shares how AI and the cloud have made a difference in genomics labs, and other areas of healthcare research like drug discovery and clinical trials.
 

Genomics


Genomic analysis can be complicated. But this is even more so in Victoria, Australia, which has a decentralised health system. Genomic study is split across multiple labs, so dispersed or duplicated data is a real concern.

The Melbourne Genomics Health Alliance, which consists of ten leading hospitals and research organisations, found a way to resolve this. It built the GenoVic platform, which allows genomic sequencing data to be shared and used across labs, clinicians and patients. The system is built on AWS cloud and is highly secure.

GenoVic is also easily scalable, so it can be used by more healthcare organisations. Five labs in Victoria are already using GenoVic to support genomic testing across different healthcare areas, including complex genetic conditions in children and adult neurological disorders.

Genomics can even be useful for studying public health outbreaks. The Genome Institute of Singapore’s (GIS) infectious diseases lab uses genomics tech to understand how certain bacteria causes diseases in humans. AWS allowed it to have a much larger capacity for analysis without having to fuss about procurement and maintenance.
 

Drug development


Developing new drugs is a costly and tedious process. The average commercial drug takes more than US$2.5 billion and at least ten years to bring to market.

US-based biopharmaceutical firm Celgene uses AI to help speed things up. In the past, researchers had to manually correct images of tens of thousands of cancer cells. Deep learning processes these images almost immediately with better results.

High-performance Amazon EC2 P3 instances have also shortened the time taken for analysis. Researchers use complex algorithms to predict how certain compounds interact with the human body. These models can take two months to train, but it is now reduced to just four hours.

This speed has given Celgene more time to experiment with new therapies and approaches. Research can move at a much faster pace to find cures for patients.

Another US biotech company, Moderna, was able to shorten its drug research and development cycles with AWS. With its IT infrastructure hosted in the AWS Cloud, Moderna received clinical trial approvals for a Zika vaccine in only a year.

AWS infrastructure helped Moderna significantly lower drug development costs, which typically run quite high. Achieving the same computing power in Moderna’s own data center would have cost millions of dollars, the company shared. In contrast, the company’s AWS infrastructure is managed by just one employee.

Moderna has invented new technologies that run on AWS to create mRNA constructs that cells recognise as if they were produced in the body. This invention has empowered Moderna to experiment rapidly on virtually any mRNA sequence, easily shifting between research priorities, without investing in new technology.

Moderna designs mRNA sequences on AWS’s highly scalable compute and storage infrastructure. Analytics and machine learning optimises those sequences for production. The company’s automated manufacturing platform then converts them into physical mRNA for testing.

In addition, Amazon Redshift – AWS’s fully managed data warehousing service – allows Moderna’s scientists and engineers to combine results from dozens of experiments that are running in parallel. They can easily share insights to refine their design and production cycle quickly.
 

Clinical trials


Once researchers have come up with a potential treatment, they need to test it thoroughly to make sure it’s safe and effective before releasing it for use. Clinical trials are an important part of this process.

But it can be tricky to get patients in clinical trials to submit high quality data regularly and securely. When patients drop out of trials or submit inaccurate data, it can lead to delays and additional costs.

UK-based uMotif’s cloud-based platform makes it easier for patients to submit e-consent or symptom data. Patients simply download an app on their own device to log the information. They no longer have to travel to a lab to do this, and researchers can now conduct trials on patients in remote locations.

The platform is built on AWS and can be used for studies with any number of participants - be it a few dozen or tens of thousands. AWS enables global hosting and language translation, so it can be used anywhere in the world.

AWS cloud is also useful for building secure and compliant systems. Pharmaceutical supply chain firm TraceLink, for instance, used AWS for a platform that helps clients comply with global track-and-trace regulations.

Healthcare researchers drive a large part of the world’s progress in medicine and treatments, and it won’t do for them to be bogged down by time-consuming processes or unnecessary costs. Tech tools such as AI in the cloud will help them in the quest to break new grounds.