The Promise of Analytics in Life Sciences

Now that I’ve described the structure of the analytics continuum in life sciences and discussed how scientists – data and otherwise – practice analytics within the space, I want to take a step back and talk about the promise of analytics and what those employing them hope to achieve in drug discovery and beyond.

In some respects, analytics are just one piece of the puzzle, but without them, solving that puzzle can take much longer, to the extent that the lives of people suffering from diseases are often hanging in the balance. Let’s explore how analytics promise to power scientific breakthroughs.

The goal of most life sciences research projects is to discover, describe, define and, ultimately, manipulate the biological processes underlying and responsible for disease states. The scientific process behind this is analogous to building and solving a jigsaw puzzle, but there’s a catch: no one knows what the final picture looks like – not yet, at least.

The strategy for solving a puzzle generally involves finding the corner and edge pieces and creating the outline – effectively bounding the problem space, like we do when starting a research project. Next, colors and visual cues on the remaining pieces allow for clumps to form within the boundaries, a process similar to the discovery of sub-processes within the biological network.

These clumps are then connected with lines of puzzle pieces in a way that is similar to the discovery of biological processes/mechanisms that relate the sub-processes in a network model. Finally, the holes are filled in with pieces that complete the puzzle, which mimics the final step of completing a complex system’s biology network related to the disease state. 

As shown above, the process of solving a jigsaw puzzle isn’t so different from the process of discovering a scientific breakthrough. In recent years, however, different practices of analytics have emerged to add a new dimension – and advantage – when it comes to producing scientific discoveries.

In the early stages of a new research effort, descriptive analytics approaches enable the discussion of new data relevant to the disease state of interest and effectively allow the problem space to be bounded in a manner analogous to the creation of the puzzle outline. Once this space is bounded, the data within it can be exploited in order to create hypotheses that can be tested using predictive analytics techniques.

These hypotheses effectively allow for the formation of clumps of related data relevant to biological sub-processes analogous to the clumps within the puzzle outline. Connections between the clumps are facilitated through the use of prescriptive analytics techniques that allow for the modeling and simulation of complex biological processes that link the clumps together mechanistically and effectively complete the puzzle for the disease state of interest. 

In an ideal setting on an ideal project, analytics promise to power scientific breakthroughs in the ways described above, but drug discovery and other scientific processes aren’t always as easy as following a linear, step-by-step process. Even so, when analytics are baked into the process and data begins to tell a very captivating story, breakthroughs are made faster and more efficiently, often saving lives as a result.

After reading this series, I hope you have a better understanding of how analytics operate within the life sciences space. If you have any questions, don’t hesitate to reach out to me directly at Christopher_Waller@EPAM.com.