What is bioinformatics?

Paulien Hogeweg and Ben Hesper described for the first time the term bioinformatics as “the study of informatics processes in biotic systems”.

In a very simplistic way, bioinformatics could be defined as a field where computational science and biology meet. This subdiscipline uses interdisciplinary knowledge from computational science, mathematics, statistics, biology and genomics to analyse large amounts of biological data and to obtain knowledge from diseases, biological processes, evolutionary events, or any other question in life science. Computational biology is a similar term and refers to applying computational methods to solve biological problems (are they the same… the answer will change depending on who you are asking). What is crucial is that computational tools are key to biology and have shaped our understanding of doing biological research.

Computational methods have sped up the advances in biology by applying techniques developed in the past (e.g. hidden Markov models, principal component analysis), and creating new ones. Moreover, algorithms can be iterated and applied (like Bayesian inference, bootstrap) to large-scale experiments at different molecular levels (e.g. proteomics, epigenomics, metagenomics, any -omic), which can be used to shed light on specific biological questions. Computational techniques have been used to study mutational processes and led to the identification of mutational signatures related to cigarette smoke and UV light, as part of a large cancer study. They can also be used to organise information on the functions of genes (e.g. the GO ontology). Combining large collections of gene expression data into a cell atlas (e.g. collecting gene expression data of thousands of cells in different tissues at different stages of development in one database) enables identification of new cell types, a better understanding of how cells communicate and the pathology of disease.

Computational methods are useful when generating hypotheses, which is circular and iterative process.. Ideas can be tested in-silico, then hypotheses can be proven experimentally and new ideas may become apparent. Some researchers go further and model biological systems to make predictions in a way to understand the biological process and combine mathematical models, computational approaches, and biology knowledge.

Bioinformatics can be used for different and diverse applications. For example, analysing the regulatory mechanisms of gene expression and drug development; analysing ancient DNA of bacteria and viruses to infer co-evolutionary relationships and diet, and analysing ancient DNA to understand migration or inferring evolutionary history by modelling population bottlenecks with polymorphisms.

In fact, the field has evolved; it is more than just analysis, it requires integration of different data (multi-omics) to create inferences, and constant interdisciplinary development to learn and apply new knowledge and methods from outside of our field – leading to the emergence of data science.

Data science involves the following ideas:

1) Systems: which includes cloud environments, workflows, benchmarking, cybersecurity;

2) Design: how to communicate between computers and humans e.g. input data, visualisation of results;

3) Analysis: interrogating the data with different techniques for example deep learning, data mining, and statistical methods (be careful not to torture the data in doing so.

4) Value: the desired step of gaining knowledge. In the end, as Bourne, P. E also said “it really doesn’t matter what you call it. Just do it”. Others argue that “computational biologists are just biologists using different tools” and we should remove this label for avoiding segregation and bringing the community closer to solving biological questions.

Comment below what you think Bioinformatics is.