Some people consider bioinformatics a service that builds databases or analyzes sequencing data. I personally see bioinformatics, or computational biology, much more broadly as a scientific research activity that involves multiple critical steps that, when poorly done or missed, will lead to sub-optimal research outcomes or even failure of the whole project.

These critical preparatory steps include the study design, interrogating novel biological hypotheses using the most relevant data resources, and determining the sufficient number of samples or time points to guarantee that the key hypotheses can be tested and research questions addressed.

When done well, these analyses allow us to then make the most of the newly-generated data to test the original hypotheses or, even better, obtain unexpected novel findings that are thrillingly common in data-driven research. Such new findings often lead to multiple iterations of follow-up analyses; for instance, machine learning or mechanistic modelling and more targeted experiments to validate model predictions. Finally, after a proper analysis of all the data, and making non-sensitive data available for others, we can together draft a manuscript with informative visualizations.

Such a multidisciplinary process requires active and continuous collaboration and communication between biological, experimental and computational researchers, whether taking place within a single group, with a bioinformatics team in-house, or as part of external collaboration. Effective communication requires mutual respect, learning from each other, open information sharing, and equal contributions and credits for all parties.

The best: team science projects

Bioinformatics is a critical part of any modern life science research that makes use of large-scale and often multi-modal profiling technologies, advanced data analysis and modelling techniques. Badly done computational biology has as equally negative an impact on the research findings as badly done experiments, and therefore both activities should be respected and credited the same way. I personally thoroughly enjoy team science projects where I can learn from the experts of the particular biological question (e.g. cancer immunology), or about a newly emerging technological approach (e.g. single-cell proteomics). 

To make such collaborative and interdisciplinary research a reality at InFLAMES, I encourage the biological and experimental researchers to make use of the skilled InFLAMES bioinformatics team and join the activities they are organizing (e.g., open-office sessions, training courses, and workshops).

Be curious and keep learning

I also encourage bioinformatics and computational biology researchers to continuously look out for and familiarize themselves with the latest technological approaches and the best data analysis practices. I myself have witnessed the evolution from microarrays to single-cell and spatial profiling, and even though the data becomes ever larger and more complex, they eventually require basically the same data analysis approaches.

As an excellent opportunity to hear the latest developments in the field of bioinformatics and computational biology, is 2024 European Conference on Computational Biology which takes place in Logomo Turku, where you will be able to see excellent examples of modern experimental-computational research in life sciences.

Until we see the emergence of the next generation scientists (or robot scientists), who will be able to do both experimental and computational analyses equally well, we must rely on mutually-rewarding collaborations to keep us at the frontline of life science research.

Tero Aittokallio

InFLAMES Visiting Professor