Insight Beyond Numbers: The Growing Significance of Visualizing Biological Data

Data Visualization is Essential to R&D

When the journal Frontiers in Bioinformatics announced a new publication venue dedicated to data visualization two years ago, it emphasized that data visualization is not about aesthetics or optional. Visualization is an essential aspect of research, necessary for data interpretation and collaborative innovation. Sean O’Donoghue, Specialty Chief Editor for Data Visualization, stated:

“Increasingly, the life sciences rely on data science, an emerging discipline in which visualization plays a critical role. Visualization is particularly important with challenging data from cutting-edge experimental techniques, such as 3D genomics, spatial transcriptomics, 3D proteomics, epiproteomics, high-throughput imaging, and metagenomics. Data visualization also plays an increasing role in how research is communicated. Some scientists still think of data visualization as optional; however, as more realize it is an essential tool for revealing insights buried in complex data, bioinformatics visualization is emerging as a subdiscipline.”

Let’s take a closer look at why visualizing biological data is so important to R&D.

Interpretation: Data visualizations help us understand complex information

In life science R&D, massive volumes of diverse data types are common. The data are noisy, complex, and interconnected, and analysis methods are intricate and recursive. While the data itself is important, how it is presented can have a profound impact. Relying on tables alone is nearly impossible; visualizations help us more easily understand complex data because our brains are wired for visual processing.

A larger portion of our sensory cortex is devoted to visual processing than to word processing. Our brains can process around 10 million bits per second and can visually recognize patterns within about 100 milliseconds. Visualizations help us better understand and retain information, making them great aids in conveying complex concepts. A recent study published in Nature Communication explained how the brain’s visual processing system is akin to a Bayesian model that processes prior knowledge with new evidence to make intelligent inferences.

Despite this, data visualization hasn’t always been a priority. Nimita Limaye in Technology Networks Biopharma notes that while real-time access to diverse sources of data is a major benefit, not enough consideration has been given to the challenges faced by end users in navigating confusing data points. However, there has been a paradigm shift from static plots to dynamic 3D visualizations, enabling deeper insights into the interplay of different parameters.

Genomics Visualization - Circos Plots

Circos plots are a good example of the importance of data visualization. They visually display complex genomics data to aid in interpreting large volumes of data. While linear plots or stacked tracks are useful for scrutinizing specific regions, researchers often need a more global view, including non-adjacent regions, to understand relationships among different data. Circos plots present complex genomics data by laying out different data tracks around a circular plot using various visual elements, such as color schemes, fonts, distances, and sizes. For example, chromosome-based circos plots can display various data points (e.g., variants, expression data, different peaks) for different chromosomes around a circle, where each segment represents a specific chromosome. With circos plots, researchers can see a more complex and complete picture of their data, helping them interpret multidimensional data such as expression patterns, somatic mutations, epigenomic profiles, structural aberrations, copy number mutations, and clinical information. Careful selection of displayed data points is essential to avoid overwhelming the analyst, and having easy-to-use tools for creating, updating, and interacting with the graphs, as well as accessing the underlying raw data, is key to interactive analysis.

The Cancer Genome Atlas - Data Visualization Tools

The Cancer Genome Atlas (TCGA) is another example highlighting the importance of data visualization in complex data interpretation. TCGA is a joint initiative between the National Cancer Institute and National Human Genome Research Institute to catalog the genetic mutations responsible for cancer. Since 2006, researchers have created and shared vast amounts of sequencing and bioinformatics data to characterize different tumor types. Over 2.5 petabytes of genomic, epigenomic, transcriptomic, and proteomic data have been created. To make sense of this data, TCGA researchers and collaborators have created numerous data visualization tools, including tools for proteomic-data visualization, cancer-genomic visualization, expression-level exploration and mutation analysis, integrated-genomics-data visualization, regulome data visualization for exploring clinical-molecular data associations, and tumor mapping. Without such tools, making sense of this volume of data would be impossible.

Innovation: Data visualizations support collaborative innovation

The high quantity and complexity of data in life sciences complicates matters. R&D organizations often have a wide range of experts using various specialty technologies to create and analyze diverse types of data. The data-science lifecycle—like the innovation lifecycle it mirrors—is an interactive process. Raw data are repeatedly created, acquired, analyzed, rendered, and observed as insights are gained and new hypotheses are pursued.

To mitigate data reproducibility issues, researchers must be able to track and share all the settings used to create data that led to meaningful discoveries, so colleagues can easily repeat their analyses. An essential underpinning to data visualization is having a data platform that can collect, integrate, and interpret data (and metadata) from diverse sources. Without it, the return on investment in gathering, organizing, collating, analyzing, and rendering data is diminished, and bottlenecks can counteract the benefits of using visualizations to quickly gain insights and make decisions.

When a scientific-data-aggregation-and-management platform is combined with powerful data visualization tools, teams gain self-service access to their data like never before. They can more easily connect diverse datasets, create dashboards and visualizations for their unique needs, and drill down into raw data. The result is the ability to quickly access data, spot trends, patterns, and hidden connections, identify outliers or knowledge gaps, and determine what questions to ask next and what avenues to explore—across many data types. Teams can view and share their data in different ways, making it easier to collaborate with other experts and, eventually, share their breakthrough discoveries. Limaye notes, “The future of data visualization is about making the process more dynamic and fueling the creative instincts of scientists by allowing them to play around with the data.”

Dotmatics Luma – Connecting Diverse Raw Data and Interactive Visualizations (COVID-19 Immunophenotyping Example)

To illustrate the importance of having a data platform that unites diverse data sources with powerful visualization tools, consider the example of COVID-19 immunophenotyping.

Early in the pandemic, a team of collaborative researchers aimed to improve the understanding and management of COVID-19 by using flow cytometry to scrutinize cell-population data from blood samples taken from highly heterogeneous COVID-19 patients who required hospital treatment. Their goal was to uncover factors linked to symptom occurrence, severity, and clinical progression, which could help guide risk-based patient-treatment protocols.

These researchers underwent a time-consuming data wrangling process to collect, model, process, and visualize vast amounts of patient and results data, which they have since made available within the Covid-IP (Covid–ImmunoPhenotype) project. If they’d had Dotmatics Luma, they could have eased some of that burden and freed time for scientific analysis. Dotmatics Luma simplifies the steps needed to transform raw data into interactive visualizations, enabling all key steps—from collecting raw data, to creating data tables and models, to outputting data for interactive visualization. Dotmatics Luma preserves the data connection from raw data all the way to rich visualizations, allowing scientists and statisticians to easily spot trends across collated and modeled data, as well as dig down into the raw data for deeper understanding.

Data Visualization with Dotmatics

As processing power increases and advanced analytics like ML/AI become the norm, data visualization will help bridge the gap between computers and humans. It will play an essential role in helping us interpret advanced-analytics results so we can more easily see what happened, understand why it happened, and predict what might come.

Dotmatics offers solutions to help better access, understand, and apply complex R&D data. In addition to their scientific-data-aggregation-and-management platform, Dotmatics Luma, they provide a range of data visualization solutions for different scientific modalities, including:

• ​Specialty research visualization: Scientific visualization tools for biologics, small molecule, and chemicals and materials research
• ​Publication-quality reporting and visualizations: Statistical analysis, graphing, and publication-quality reporting with GraphPad Prism
• ​Decision-support visualizations: Advanced data visualization and analysis for decision support with Dotmatics Vortex

References

1. 1.O'Donoghue, S. Grand Challenges in Bioinformatics Data Visualization. Front. Bioinform., 17 June 2021. https://doi.org/10.3389/fbinf.2021.669186
2. 2.Kouyoumdjian, H. Learning Through Visuals - Visual imagery in the classroom. Psychology Today. July 20, 2012.
3. 3.Koch, K., McLean, J., Segev, R., Freed, M. A., Berry, M. J., Balasubramanian, V., et al. (2006). How Much the Eye Tells the Brain. Curr. Biol. 16, 1428–1434. doi:10.1016/j.cub.2006.05.056
4. 4.Healey, C. G., and Enns, J. T. (2012). Attention and Visual Memory in Visualization and Computer Graphics. IEEE Trans. Vis. Comput. Graph. 18, 1170–1188. doi:10.1109/TVCG.2011.127
5. 5.Evolution wired human brains to act like supercomputers. ScienceDaily. September 14, 2023.
6. 6.Harrison, W.J., Bays, P.M. & Rideaux, R. Neural tuning instantiates prior expectations in the human visual system. Nat Commun 14, 5320 (2023). https://doi.org/10.1038/s41467-023-41027-w
7. 7.Limaye, N. Data Visualization in Biopharma: Leveraging AI, VR and MR to Support Drug Discovery. Technology Networks Biopharma. June 12, 2019.
8. 8.Krzywinski M., Schein J., Birol I., et. al. Circos: an information aesthetic for comparative genomics. Genome Res. 2009 Sep;19(9):1639-45. doi: 10.1101/gr.092759.109.
9. 9.Nattestad, M. Making genomic data come alive with circos plots. Medium. September 25, 2017.
10. 10.Visualizing Genome: Creating Circos Plots. Bioinformatics & Research Computing. Whitehead Institute.
11. 11.The Cancer Genome Atlas Program (TCGA). National Cancer Institute - Center for Cancer Genomics.
12. 12.TCGA Computational Tools. National Cancer Institute - Center for Cancer Genomics.
13. 13.Hoadley, K.A., Yau, C., Hinoue, T. et al. Cell-of-Origin Patterns Dominate the Molecular Classification of 10,000 Tumors from 33 Types of Cancer. Cell. 173 (2), 2018. https://doi.org/10.1016/j.cell.2018.03.022
14. 14.Laing A.G., Lorenc A., Del Molino Del Barrio I., et. al. A dynamic COVID-19 immune signature includes associations with poor prognosis. Nat Med. 2020 Oct;26(10):1623-1635. doi: 10.1038/s41591-020-1038-6.
15. 15.Hayday, A., Edgeworth, J., Shankar-Hari, M. Covid–ImmunoPhenotyping - a preliminary data release. COVIDIP - Infection, Immunity, Immunophenotyping. May 22, 2020