Mainstream Precision Medicine Hinges on Healthcare


Adam Marko, Director of Life Science Solutions at Panasas

The healthcare industry is experiencing a data deluge, and it shows no sign of slowing down.

In 2020, estimates reported that healthcare data comprised a staggering 30% of the world’s total data volume; projections for 2025 bump that number up to 36%, a growth rate outpacing every other industry.

This is promising news for precision medicine, the data-driven healthcare initiative that promotes the right treatment for the right patient at the right time.

Precision (or personalized) medicine revolves around the understanding that medical conditions and drug responses vary from person to person – for example, chemotherapy that shrinks one patient’s tumor might be fatal to another. It combines novel tools from the omics sciences (genomics, pharmacogenomics, and proteomics), advanced medical imaging techniques, and artificial intelligence (AI) to tailor diagnosis and treatment to the individual patient.

While the approach currently sees most action in oncology, the innovations required to extend it to all levels of care are already underway. But to realize the dream of mainstream precision medicine, the healthcare industry must first reckon with the intensive compute requirements it entails.

We should therefore expect to see the implementation of high-performance computing (HPC) infrastructures capable of integrating all the necessary biomedical and health data into clinical workflows. Given the massive volume of that data and its exponential growth, healthcare organizations will ultimately need to invest in cost-effective, easy-to-manage storage solutions that are fast, scalable, and secure.

“What brings you in today?”

You probably heard this greeting at your last primary care visit. In response, you rattled off a list of symptoms; maybe you even forgot a few, or you didn’t think they were worth mentioning.

At any rate, your self-report formed the primary basis of your doctor’s diagnosis, and she ordered the tests necessary to confirm it. Maybe she was right and selected a successful treatment. Or maybe she was wrong, and you found yourself back at square one with another appointment booked for next month.

This is the traditional approach to healthcare: It is reactive, generic, and resigned to trial-and-error diagnoses and treatments.

In a future where precision medicine has become the standard model of care, your doctor is armed with your individual “dataome,” your body’s biological blueprint combined with your total medical history. This would include your genetic profile, unique protein signatures, behavioral and environmental factors, all the details from your electronic medical records, and the information stored on any wearable medical monitoring devices.

That treasure trove of data will have been analyzed not just by your doctor, but also by advanced AI algorithms capable of cross-referencing it against massive databases and drawing insights about you that were previously impossible.

Making this data accessible to doctors will not only save money and time, but also lives.

For instance, most people don’t know that pharmacogenomic research has determined that over 200 commonly prescribed drugs metabolize differently depending on patient genetics. Even more shocking still, almost everyone has at least one of the genetic variants that interact with those medications. One of those could make you so sensitive to a drug that you only need half the standard dosage; another could mean that you require twice as much; yet another could produce a lethal reaction. A DNA test at the doctor’s office screening for those variants, performed before your prescription is ever written, could prevent one of the yearly 1.3 million emergency room visits in the US alone that are linked to adverse drug effects.

This is the power of precision medicine in action: Utilizing all the data available to make the most informed decisions for the best patient outcomes.

We’re already on the path

HPC is powering the scientific breakthroughs that get us closer to this goal.

Consider the rapid growth of genomics. When scientists set out to sequence a whole genome for the first time in 1990, it took them 13 years to do it. Today, researchers at Stanford Medical just set the new Guinness World Record for the fastest DNA sequencing, clocking in at five hours and two minutes. Less than three hours after that, they used those results to successfully identify a genetic disease. This shattered the previous 14-hour record for a genome sequencing-based diagnosis.

Thanks to advancements in sequencing technology supported by HPC, what once took over a decade and cost nearly $ 3 billion is being increasingly done in less than a workday and for less than one thousand dollars. The plummeting timeframes and costs mean that we are rapidly approaching the possibility of practicing genomic medicine at the primary point of care.

Proteomics, the field that investigates healthy and diseased cell processes at the protein level, is experiencing similar advancements. Because protein signatures can explain the link between genotype and phenotype, proteomics complements genomics and plays a key role in advancing precision medicine. Currently, a team of scientists from the Department of Energy and the Georgia Institute of Technology is combining HPC and deep learning to predict the structures and functions of thousands of proteins. Their findings will put even more critical data into doctors’ hands.

Beyond the omics, HPC is also transforming the realm of diagnostic imaging through AI applications that serve the interests of precision medicine. For example, scientists are deploying AI to identify image biomarkers that can help doctors make better-informed treatment decisions, preventing unnecessary biopsies and excessive radiation for cancer patients.

We need Healthcare HPC to harness the data deluge

Mainstream precision medicine will mean faster and more accurate diagnoses, safer and more effective treatments, and early disease detection and prevention – for everyone, everywhere. But it depends on the ability of healthcare providers to derive meaning from the data deluge.

For big data to become clinically actionable data, modern healthcare organizations will need to follow the lead of the researchers and implement HPC storage solutions that are up to the task. Making the investment will revolutionize health as we know it.

About Adam Marko

Adam Marko is the Director of Life Science Solutions for Panasas. In his role, Marko is involved in driving all aspects of market development in Life Sciences including working with field sales, marketing and engineering. He has more than 15 years of experience as both a researcher and an IT professional, analyzing data and meeting the informatics needs of life sciences organizations. Mark can be reached at [email protected].

Source link