Does big data and TDA hold the key to accelerating drug discovery?

pharmaphorum's Hannah Blake recently had a chance to speak with Pek Lum, VP of Solutions and Chief Data Scientist at Ayasdi, to discuss why Topological Data Analysis (TDA) may soon revolutionize the drug discovery process.

"Big data" is a term that, in just the last two to three years, has become part of our common lexicon. According to reports in both the mainstream press and industry trade publications alike, big data promises to reshape such diverse fields as financial services, marketing, national security, and the life sciences. Certainly, our industry is no exception. Pharma companies of all sizes are busy analyzing their vast troves of data in an effort develop their next breakthrough product.

With that in mind, pharmaphorum's Hannah Blake welcomed the opportunity to speak with Dr Pek Lum, to learn about Topological Data Analysis, otherwise known as "TDA". TDA is a specialized form of data analysis that is currently being used by researchers at drug makers and academic institutions – such as Merck, BN ImmunoTherapeutics, UCSF and Mt. Sinai – to identify potential cures for a wide ranging set of disorders and disease such as cancer, autism and post-traumatic stress disorder.

Dr Lum, a biologist with an expertise in molecular genetics and cell biology, shares with us her thoughts on why TDA holds promise for big pharma, how it differs from other sub-sets of big data analysis, and why drug development is no longer a "chemistry problem".

Interview summary

HB: Hello Dr Lum, it's great to speak with you. Why is drug development no longer a chemistry problem?

PL: Finding the "one cure" for a disease is almost impossible because both diseases and drug responses are incredibly complex. For pharmaceutical companies, that means making a drug is no longer the issue. Making the right drug for the right sub-population is the real problem. Pharma is tackling this challenge head on by generating vast amounts of genomic, genetic, chemical, and clinical data. Because these datasets are extraordinarily complex, conventional analyses are unable to scale. The 'wet work' is usually the most time consuming phase of the drug development process. Now, it is the data analysis phase that demands the most time, resources, and expense.

HB: What is Topological Data Analysis (TDA)?

PL: TDA is an applied mathematical approach to characterize the shape of data. We like to say that data has a shape and that shape has meaning. Using machine learning algorithms and statistical functions, TDA can be applied to complex data and automatically find relevant patterns and groupings. For pharma, TDA enables subject matter experts, computational biologists, chemists, and clinicians, to interact directly with data and automatically discover previously hidden insights – sometimes in just seconds.

HB: How do pharma companies benefit from this new way of analyzing data?

PL: Topological Data Analysis (TDA) allows researchers to identify patient sub-populations, to identify biomarkers and to understand their underlying etiologies. Researchers are able to go through data faster and turnaround analyses and reports without relying on IT support to script queries. With a stronger approach to analytics, pharma companies and researchers can drastically shorten the time it takes to bring a drug to market. Not only does this bring massive cost savings, but it also brings better medicine to people.

HB: What are the biggest challenges for pharma using TDA?

PL: Adopting a new technology is always a leap of faith. It requires both the desire to explore new innovations and the fortitude to risk trying something new. TDA makes it possible for a pharma to move more quickly through clinical trials and to decode large amounts of data generated from RNA and DNA sequencing. Some of the largest and most well respected life sciences organizations are already finding new insights using TDA, such as the FDA, five of the top fifteen pharmaceutical companies including Merck, Mount Sinai, UCSF, and Harvard Medical School.

HB: What difference can TDA make in the drug discovery process?

PL: TDA has the potential to dramatically impact how pharmaceutical companies formulate drugs right now. First, TDA is inherently suited for tackling complex data such as those found in drug discovery, because it uncovers signals that are hard for other methods to find. Second, by transforming biologists into data scientists, TDA can dramatically decrease the time it takes to find insights and correlate complex genomic, chemical, and clinical data. What had taken months can now be reduced to a few hours. We envision that with TDA, a pharma company can take a drug to market in as little as 4-5 years.

HB: What other breakthroughs in pharma do you see happening in 3-5 years with TDA?

PL: Some of our primary goals are to enable pharma companies to identify patient sub-populations, tailor medicines for those specific subgroups, and to increase efficacy and decrease adverse events. By fully leveraging and integrating data from next-generation sequencing and clinical data during the next few years, groundbreaking insights and advances will be found in autism, cancer, Parkinson's disease, Post Traumatic Stress Disorder (PTSD) and Traumatic Brain Injuries (TBI). As we continue to make technological advances and empower people across entire organizations to new insights in data, we will begin to successfully manage and cure some of the most confounding, harmful, and expensive diseases that we face today.

About the interviewee:

Pek Lum leads Ayasdi's products and solutions team. After more than a decade in the life sciences industry, Pek has the passion to help find a cure to cancer and other diseases in her lifetime, and believes that Ayasdi's technology may help solve these difficult problems. Pek was trained in molecular genetics and cell biology for her Ph.D. at the University of Washington. Her work has been widely published in scientific and medical journals, and her research has contributed to discoveries in drug development and the understanding of complex diseases. Pek spent 10 years at Rosetta Inpharmatics and was part of Merck's $620M acquisition of Rosetta in 2001, one of the companies that in many circles is known to have started the genetics revolution with its bioinformatics solution for deciphering enormous amounts of gene-expression data to derive insights into numerous potential therapeutic targets for drug discovery. Pek also has a Master of Science in Biochemistry from Hokkaido University in Japan, speaks more than 5 languages, and lives in Palo Alto with her husband and two children.

Why is drug development more about effective use of data than chemistry?