How big data can drive innovation in drug development and personalised medicine

11 November 2015

Tom Edwards

Tom Edwards

Former Consultant

Big data has become increasingly important in driving performance in many industries. The Pharmaceutical industry is no exception. Medicine is shifting from a curative to preventative environment by combining clinical and genomic data and is being enabled by advancements in data processing.

This has led to the cost of mapping a human genome being set to be less than $100 by 2016. This process was started twenty-five years ago when scientists began a long-term project in the UK to sample and analyse data trends to discover new links between living conditions and health implications.

Children of the 90s

In the early 1990s, a ground-breaking scientific study began to take place. Almost 15 thousand pregnant women took part in a study called Children of the 90s to track them and their children over the coming years. The study surpassed all expectations and still continues today, nearly 25 years on. Over 500 of the original children are now parents themselves and their children are being tested and tracked. This has resulted in the project becoming the first to ever collate substantial data from three generations.

Over the years, each participant has had to fill in regular questionnaires and occasionally visit a lab for tests. By analysing these vast amounts of data, scientists have been able to see trends and make scientific discoveries. These include:

  • People with two copies of a particular gene variant are 70% more likely to become obese
  • Eating oily fish during pregnancy may improve children’s sight, IQ and social behaviour.
  • There is a link between peanut oil and peanut allergy which has resulted in all skin creams having to list ingredients

Falling costs of genomic sequencing

While this project was ongoing, the first ever human genome was mapped as the conclusion of the Human Genome Project in the year 2000. The project took ten years of work and about $200m just to map a single person’s genomic makeup.

Since then, as shown by the graph below, big data analytics have driven down costs dramatically over the last fifteen years. As of this year, the cost of mapping a human genome is under $1000 and is set to reduce by a further factor of ten in 2016.

 


 

Looking ahead

For ‘Children of the 90s’, scientists were able to see trends in people’s lifestyles compared to their children and grandchildren. When the Human Genome Project was in its infancy, scientists could see further patterns and make broad predictions about human disease based on a small sample set.

As we are soon to be in a position where the 1.5 gigabytes of a human genome can be mapped for $100, we can combine genomic data with clinical data to gain better insights into a person’s makeup. This is likely to result in an increase in personalised medicine as we can predict a patient’s responsiveness to drugs and risks to particular diseases by analysing their genome – shifting medicine from curative to preventative.

This can be enabled by using a real-time Clinical Data Warehouse based on the SAP HANA in-memory database. This can be used as a basis to analyse data sets in real-time for bio-statistical and explorative analysis as well as predictive analytics. Patient subgroups with similar genetic makeup, lifestyle, or medication and treatment history, can therefore be identified more easily which can help find out key influencing factors of the effectiveness of a new drug.

This wealth of new data and improving analytical techniques is set to enhance future innovation of drug development and personalised medicine.

Bluefin and SAP S/4HANA - welcome to the one horse race