Information is ubiquitous, the expanse of which is so vast that the present age is historically termed as “The age of information”. Nevertheless, another pragmatic statement by American military historian Caleb Carr, “It is the greatest truth of our age: Information is not knowledge.” holds true to its very core. The intervention of a new-found stream of science, the Data Science, plays a massive role in an utilitarian transfer of information into a more plausive knowledge of not only what has happened in the past but also into what will happen in the future. The goliath of the mid-to-late 20th century, the pharmaceutical industry, also could not avert the slingshot of this calibre of information.
So, what is BIG DATA?
The science has seen a fascinating change lately but wisely—transition of coining the jargons from Latin to colloquialetopian English—Big Data, therefore, is nothing more than a large volume of relevant information. It is not exactly about what a small group of people posting about a health condition on Facebook or searching for somedespicable remedy of a terrifying disease on Google. This focuses on what theimprints, the mass is leaving behind the pan-continent. The relevant piece of information thus can infer whether you shall be able to repay your heath loan, or which medical college you might take admission in, or whether you are susceptible to a severe mental disorder. Such ‘predictions’ are not any shamanic tricks, but the exemplifying effort of a data scientist.
The influx of such data can be varied. It can be collected from your recent hospital visit, your buying habit, the medicine you are taking recently. In a structured method, the same influx can be correlated to the quality of a raw material in sampling, the production data of bulk material, errors in packaging, apparently irrelevant data that deemed the scientist’s hypothesis unsuccessful and even the economic condition that governs the ability of a patient to buy a branded drug or its generic alternative.
The highest influence of Big Data in pharmaceutical sector—R&D or Market research?
There is no doubt that Data science can influence a lot in the sector, however, without haranguing on the meticulous details of each of the field, one must focus on the segment, where data can influence the most:
- Market research: Market research is crucial for a new product launch. A keen analyst’s eye can capture the behaviour of the customer (i.e the health care professionals including doctors and nurses) and the end user (the patients). This, in conjunction with demographic and para-economical data of a localised bunch, can reduce the cost of prodigal ad campaigns, optimise the profit margin and fine tune the logistic support.
- Research and Development: Like any other industry driven R&D sector, several dataset of present disease scenarios, can lead to identification of the conundrums of health-care systemsbeing manifested. If a disease is far more prevalent that the other, a focused team of research scientist can not only make a break-through but also be highly rewarding the industry itself. Even if we consider core research works such as drug discovery, big data related tools such as library screening, drug repurposing, natural product library, drug-receptor interaction database are making research more parsimonious compared to what it used be a few decades back.
- Clinical research: As a highly regulated sector, the clinical research field in an exemplar of the influence of data science. The various stages of clinical trial require a myriad of data from diversifiedvolunteers and a swift and accurate inference of the pattern demonstrated through statistical modelling. Even post market surveillance and pharmacovigilance has shown significant result in identifying the egregious drugs which were initially thought of being apparently safe.One such major example is the drug Nimesulide which was initially regarded as an equally innocuous and effective alternative to Paracetamol, later showed a great amount of hepatotoxic adverse effects. Undoubtedly, data analysis bolstered by tireless collection of post marketing safety data led to a greater safety for the patient.
The above-mentioned points can only gently brush the tip of the iceberg. If we consider the health care sector at large, the rabbit hole runs deeper. Big data amalgamated with machine learning results into artificial intelligence. This artificial intelligence can minimise the error in prediction, fastens diagnosis for a miniscule financial cost. In the ongoing pandemic as we face at the time of writing this article, epidemiologists of the world are relying on cold hard numbers to morph them into a fruitful prediction to successfully evade the chaos and rejuvenate the nose-diving economy.
To keep up the pace—upgrading yourself:
Of course, as any professional enthusiast, big data opens a scope of unfathomable enormity and keeping a dead eye will be ingenuous. For any pursuing pharmacy graduate, a few upgrades will be unavoidable.
- Brush up your statistics and maths: Gone are those days where the subject
- was considered to be related to biology and chemistry. The basic knowledge of maths and statistics will give unorganised numbers to a more manageable data. Visualisation relaxes the nerves by making those data even more plausible ones. Start form the considerably basic like the mean, median mode and grow to more complex modelling and curve fitting.
- Learn to handle the data: You will not want your tiresome domino art to fall before a complete set up. It is extremely important to learn how to tackle the colossal amount of data and prune them. While advanced knowledge of MS Excel can be beneficial, more concentrated tools such as SQL, Oracle will be proven functional at a later stage. However, a basic idea will be good at an initial phase which can be honed as per the future requirement.
- Fall in love with computer, in its true sense: A PC is an all-round entertainment device, but for a data scientist, use of can be a lot more object oriented. The tools for data computation and statistical handling can be broadly classified into two segments—tools with GUI and tools with CLI. Graphical User Interface (GUI) tools such as Minitab, Staistica, SAS, SAP are easy to operate as they are mostly based on visual cues. The Command Line Interface (CLI) programs such as python, R, J, Rubi, are difficult to grasp but shall be more rewarding in terms of flexibility. Both subcategories of tools have growing libraries to tackle difficult conditions.
- Mechanistic approaches—putting the common sense into work: All these computations will result into some mathematical equation and to some extent, a majestic graph with beautiful colours. However, to convert the art into a deductible science human intelligence and experience is of utmost importance. This is known as the mechanistic representation of a model. Reading through relevant papers, following the work of data scientist and thorough perusal of case studies will not only gather relevant experience but hone the deduction capabilities as a budding data scientist.
Amidst the proliferation of real-time and historical data from sources, such as, web, social media, sensors, log files and transactional applications, Big Data is rapidly gaining traction from a diverse range of vertical sectors. The healthcare and pharmaceutical industry is none other the exception to this trend, where Big Data has found a host of applications ranging from drug discovery and precision medicine to clinical support and population health management. In this data frenzy time and age, strong analytical prodigue infrastructures are leading the pharmaceutical sector. Data democratization is the next step in pharma analytics and is essential during today’s ruthless markets. The world is changing at a rapid pace. May be even a lot more than we can imagine right now. May be a new hypothesis is being born from data while we finish reading this article. While big data and AI might threaten the existence of menial jobs, highly skilled professionals shall have the tenacity to reign over these wild horses and ride to places limited by the horizon itself. There was a saying, “He who holds knowledge shall have the power”, now it’s surely, he who can manage data shall have the power.
