Exploring the Potential of Big Data in Biotechnology

By Stefano Tommasone Ph.D.Reviewed by Megan Craig, M.Sc.

With continuous advances in new experimental techniques, such as sequencing and high‐throughput technologies, biotechnology is becoming increasingly dependent on using big data to develop, test, and commercialize new or improved products and services.

Over the last decade, new technologies with faster and less expensive methods have enabled us to perform an increasing number of experiments at a rate that was unthinkable just a few years ago. These experiments generate large amounts of data, which are stored in thousands of databases, both public and private.

Nowadays, it is almost impossible to make significant advances in biotechnology research without using approaches based on big data. Automated tools are necessary to integrate data from different sources and process and exploit the information available to generate new knowledge.

However, there should be some caution. There are challenges concerning how large and complex data sets are managed and processed. Ensuring the quality, reliability, and validity of the data is crucial and requires rigorous testing, evaluation, and verification.

Having measures in place to protect the security and privacy of the data and related applications is also another important aspect.  

What is Big Data?

Big data is data so large, diverse and complex that it cannot be managed with traditional processing methods. New architectures, techniques, algorithms, and analytics are required to process and extract value and knowledge from such data sets.

The data (texts, images, signals, metadata, etc.) can be unstructured and structured. It may also refer to different dimensions and time scales. Hence, in addition to the challenge of managing high volumes, there is a lot of complexity linked to the variety of data and its proper management and exploitation.

The two main approaches used to explore big data in biotechnology are based on supervised and unsupervised machine learning, respectively. A common unsupervised machine learning method is clustering, where sets of data (unlabeled instances) are grouped by similarity according to their properties.

Supervised methods instead use labeled datasets to train algorithms to classify data or predict outcomes accurately and find more applications in biotechnology. From regression and decision trees to deep neural networks, many techniques can be used to perform supervised machine learning.

The Impact of Big Data in Biotechnology

High-throughput analysis generates terabytes of data. There are currently thousands of databases specific to biotechnology that include information about human genes and diseases, protein sequences and structures, and metabolic and signaling pathways, to name a few.

Relevant information is also derived from electronic health records and other sources of medical data. Public health services across the world have started storing and analyzing information on hospital admissions, drug prescriptions, and specialist visits.

Using big data can help identify patterns, trends, and correlations that might otherwise be overlooked or hidden with conventional methods. It also enables to test and validate hypotheses faster and more accurately. For instance, big data can assist in designing and optimizing new drugs, identify potential drug targets, and reduce errors in clinical trials.

What is Big Data Used For?

Drug discovery, genomics, precision medicine, and disease management provide some examples that show how biotechnology research can be supported by big data.

Drug discovery is probably one of the most promising areas. Using classifiers trained on compounds with known properties can help determine the behavior of complex molecules. It is also possible to create virtual assays to identify potential drug candidates that can then be tested in the laboratory.

When using the appropriate set of properties, it is possible to use classifiers trained on other molecules to identify whether a new compound is able to affect a given pathway or eliminate a specific pathogen.

In genomics, such as genome-wide association studies (GWAS), machine learning and big data can be used to deduce genotype‐phenotype associations and identify the relations between genetic characteristics and the response to specific treatments. This can lead to applications in personalized medicine and cancer research.

Following the well-known Human Genome Project, which marked a major milestone in the field of genomics, many research projects have been launched over the last two decades aimed at collecting and making available large amounts of individual genome information, including the 1000 Genomes Project (2008), the Million Genomes Project (2018).

Video Credit: TED/YouTube.com

Current Issues with Big Data

There are some important issues that should be taken into account, specifically concerning the reproducibility of the results. Several publications in scientific journals have been retracted due to a lack of reproducibility caused by a mislabeling of the data.

There are also implications related to privacy and data access. Population-based studies can unveil correlations between human health behaviors and common diseases or enable individual-level studies involving phenotype, genotype and exposure data. Since there is a risk that data may be used for purposes different from the original one, proper actions should be taken to deal with the issue.

Future Perspectives

Big data can help push the boundaries of biotechnology research and maximize its impact. Particularly, with the rapidly increasing number of experiments and the huge volumes of data generated, big data will become instrumental.

It is crucial that scientists develop the skills and methods needed to deal with the data. Attention should also be paid to the potential issues associated with lack of reproducibility, as well as privacy and data access management.

Further Reading

• All Biotechnology Content
• What is Biotechnology?
• Dissecting the Current Landscape of Biotechnology in Europe
• The Colors of Biotechnology; What do they mean?
• Importance of Biotechnology in Agriculture