Page 110 - Building Big Data Applications

106 Building Big Data Applications

can be made feasible with seamless collaboration between the industry, academia, the
regulators, governments, and healthcare providers. But the issue here is there is too
much data floating around, not all of it is used, we still struggle with integration of
technology to answer the foundational problems. There is a solution with big data
platforms, and we have seen several leading pharmaceuticals unraveling the discoveries
made with the use of big data infrastructure.
To understand how the big data platforms become useful, we need to take a peek
into history. In the early 1990s there was a definition called as Eroom’s law, which
primarily is the observation that drug discovery was becoming slower and more
expensive over time. This slowdown was happening despite improvements in technology
including high throughput screening, biotechnology and chemistry, and computational
drug-design. The cost of developing a new drug roughly doubles every 9 years with all
inflation adjustments was the definition outcome from the observation. The name of the
law was the reverse of Moore’s law in technology, to show the issue. There are significant
areas that led to the observation being formed and these included the following:

Minimal incremental effect: an opinion that new drugs only have modest incre-
mental benefit over drugs already widely considered as successful, and treatment
effects on top of already effective treatments are smaller than treatment effects
versus placebo. The smaller size of these treatment effects mandates an increase in
clinical trial sizes to show the same level of efficacy, which results in longer delays
and inefficiencies that were not integrable until big data application, mashups, and
platforms became a reality.
A cautious regulator issue: the progressive lowering of risk tolerance seen by drug
regulatory agencies makes drug research and discovery both costlier and harder.
After older drugs are removed from the market due to safety reasons, the bar on
safety for new drugs is increased, which makes it more expensive for the process to
complete and this causes a slowdown.
Spend more and get where? the tendency to add human resources and system re-
sources to R&D, which may lead to project overrun. This is a very precarious situa-
tion where you are at a point of no return. This is eliminated in the new landscape
to a large extent.
The brute force bias: the tendency to overestimate the ability of advances in basic
research and brute force screening methods to show a molecule as safe and effec-
tive in clinical trials. From the 1960s to the 1990s (and later), drug discovery has
shifted from whole-animal pharmacology testing methods to reverse pharmacology
target-approaches that result in the discovery of drugs that may tightly bind with
high-affinity to target proteins, but which still often fail in clinical trials due to an
underappreciation of the complexity of the whole organism. This issue is a brute
force technique which provides limited success if at all and often has led to more
failure.