Page 112 - Building Big Data Applications

108 Building Big Data Applications

Minimal animal testing and dependency
Refine the trials list to the most viable targets
Real world data (RWD)
Deploy in cloud
Design patient diary for tracking and compliance
Electronic medical record (EMR)
By creating a series of data files with JSON or XML, we will bring data from these
subject areas into the big data infrastructure. Our next goals are to create a series of
predictive models and outcomes which will plot the expected results from each exper-
iment and outcome. This model will accelerate the overall sharing and collaboration of
data and experiments which will be used by multiple teams and their associated ex-
periments. This type of data collaboration will provide us a pace of innovation that
benefits us as a community. How did we get this far? The ideation is to have a file
systemebased data platform that can be leveraged to store large and small data files, all
these files will contain data from drugs, patients, diseases, experiments, predictive
outcomes, actual outcomes, expected end results, community predictions, community
anticipations, community classification of disease stages, community response, com-
munity social media outreach, government agencies inspection, compliance and regu-
latory filings, actual patient reports and filings, actual providers research data, and actual
overall outcomes. All of these mean that we still do drug discovery, but the process
optimizes and transforms like this.
The process model is followed based on specific research and drug outcomes as
desired by the pharmaceutical. The goal here is to beat the disease state and discover
cure that can help the management of the disease. We have progressed with more drug
discovery in the last few years based on genomic treatment with a patient specific state
of treatment based on their gene and the reaction to the drug. This work is done once the
gene is extracted from the patient and then research is driven in the lab, and outcomes
are studied based on several factors, with the underlying behavior driven by the protein,
RNA, DNA, and other molecular structures.
The continued success we are seeing today is aided with machine learning, neural
networks, artificial intelligence, molecular research, bioscience, statistics, and continues
to improve and generate global collaboration.
In all of these processes discussed so far, now bring in the complexity management
and look at the processes and steps described. We can easily implement the manage-
ment of complexity from acquisition of data, through ingestion, integration, segmen-
tation, classification, analysis, operational analytics, predictive analytics, data lake
transformations, data hub transformations, analytics, and visualizations. Most of the
industries in pharmaceutical area today have a combination of heterogenous technol-
ogies and they all have made vast improvements in data strategy and management. With
the evolution of cloud computing and microservices architecture, we are seeing more
improvements in the reduction of complexity.