Currently, bioinformatics is playing an increasingly important role in life science research. Biologists, clinicians and biomedical researchers have become more dependent on bioinformatics outcomes. Despite the crucial role of bioinformatics in accomplishing multidisciplinary projects, collaborations between biologists and bioinformaticians encounter several difficulties. Here, I outline different types of collaborations and provide an overview of how the relationship between bioinformatics and life science experts can be facilitated.
Potential means of collaboration
Several options for collaboration are available to research groups. Depending on resource availability, they can hire their own bioinformaticians, collaborate with a bioinformatics group within their organization, use the services of an internal bioinformatics core facility, or employ a bioinformatics consulting company outside their organization.
If a bioinformatician is hired by a given research group, all members of the group can have easy access to bioinformatics assistance, communication is easy, and both life scientists and the bioinformatician will feel they have ownership and input into the project, so conflict over assigning credit for output is less likely. However, the work environment might be less competitive for bioinformaticians hired for a specific task because there is insufficient opportunity to expand their expertise, given that they focus on certain types of data related to a specific topic, and there may be minimal constructive criticism from colleagues.
If a research group decides to collaborate with an internal bioinformatics group, creating a good working relationship is more challenging. Bioinformatics groups within academia typically have their own research projects, so their ability to assign time and services to other groups may be limited. Consequently, it is important that resource allocation and expectations should be clearly established and both sides should agree on anticipated credit gains. The risk for bioinformaticians in this scenario is that despite devoting time and expertise to external projects, their contributions might not be valued sufficiently. For example, despite a bioinformatician providing data/results of publishable quality, they may not be assigned a correspondingly prominent position in the authorship list of the resulting scientific papers.
Nowadays, many research institutes have established their own bioinformatics core facilities, with the objective of supporting all research groups in the institute. This could represent an ideal model of collaboration provided that the core facility has a sufficient number of experts, expertise, and resources to tackle the research questions it is presented with. In reality, given the considerable diversity of life science data, newly established core facilities are unlikely to have a sufficient depth of experience to handle all types of data. Accordingly, research groups should be aware of such limitations and must be willing to help core facility bioinformaticians to develop their skills. Where such core facilities are overburdened with requests, there might also be a significant delay in data analysis and revisions.
Finally, outsourcing data analysis by recruiting the services of a professional bioinformatics consultancy is another option, but it appears to be adopted less in academic contexts. Compared to bioinformatics core facilities in academia, professional consultancies tend to be better at project management and generally do not expect authorship rights in publications. However, they are likely to be more expensive, accessibility might be more limited compared to the previous options, and the limitations of core facilities can also be relevant to external consultancies.
Who should sit where?
Bioinformatics is many things. As an interdisciplinary field of science, it has multiple applications including database creation and management, development of software and analytical tools, creation and implementation of computational pipelines to analyze next generation sequencing data, gene expression studies, prediction of macromolecular 3D structures, drug design, precision medicine, phylogenetic studies, amongst many others.
This multitude of applications means that bioinformaticians also tend to have different specialties. It is relatively rare to find a bioinformatician that possesses experience in all or even many of these applications. Life science data is diverse, expansive and complex. Mining such “big data” to extract useful knowledge is complicated and requires careful analysis using appropriate techniques. Mistakenly, bioinformaticians might be seen as “a jack of all trades” by some life scientists, who may think that a bioinformatician should be able to do all types of analysis quickly just by running a few lines of code.
To achieve a successful collaboration, it is crucial that all contributory parties clearly establish the goals, requirements and scope of the project, allowing the right person(s) to be recruited for the right task. For example, if a specific algorithm or computational tool must be developed for a project, it would be more relevant to recruit a bioinformatician with a computer engineering background who can rapidly develop the desired tool. Alternatively, if assistance in data analysis is needed to answer a specific biological question, then it would be better to recruit a bioinformatician with a biological background, who could better comprehend the research context and apply or modify appropriate tools and pipelines to fulfill the needs of the research group. Since biological applied research often involves several rounds of data analyses, data optimization based on feedback, and repetition of pipelines on different datasets, strong lines of communication are essential.
Similar principles should be considered when selecting the leaders/coordinators to manage multidisciplinary projects. A bioinformatics leader should be familiar with the challenges of a broad diversity of bioinformatics applications. He or she should be acutely aware that applied works are as challenging as development tasks and that sufficient time and resources should be allocated to teaching bioinformatics to biologists. It is crucial to understand the needs of life science researchers and to plan resources accordingly so that those needs can be met. The leader of a bioinformatics group should also ensure that the right person(s) is assigned to each project and that whoever requested bioinformatics help is comfortable with the person and process allocated to them.
Assignment of credit
Appropriate assignment of credit is another important factor to maintaining a high level of motivation in collaborations between life science and bioinformatics experts. Credit should be distributed fairly between those who own the scientific idea, those who produce the primary data, and those who add value to it through data analysis or the development of analytical tools. Assigning credit in multidisciplinary projects is a relative concept, and it can be a significant source of conflict, being very much dependent on the characteristics, scope and contributors of a project.
If development of algorithms and computational pipelines is the main focus of the project, most of the credit is attributed to the bioinformaticians whereas, in applied works, partitioning of credit can be more challenging because measuring added value and comparing it among contributors is difficult. Since the life sciences largely remain the domain of biologists, there might be a risk for bioinformaticians to be viewed more of as service providers rather than scientific partners.
Conducting multidisciplinary projects is challenging and success requires a coordinated effort by all contributing disciplines. To facilitate the cooperation necessary between bioinformaticians and life scientists, firstly, it is important to bear in mind that the life sciences and bioinformatics are dependent on each other. Without bioinformatics it would be impossible to manage and analyze the ever-growing amounts of data from life science research and, without that “big data”, bioinformatics could not gain its prestige.
Secondly, human resources have a central role in creating the good working relationships necessary to enable successful collaborations. It is crucial to find a suitable bioinformatician for each role, to be clear about expectations, to provide opportunities for skill development, and to listen to feedback, all of which will help ensure that good bioinformaticians are retained. Managers have a very important role in facilitating collaborations, and it is their responsibility to create an environment that bolsters employee satisfaction because “people leave managers, not companies”.
Thirdly, the needs, interests and benefits for both sides of a collaboration should be well aligned. Only when everything is based on mutual advantage can optimal performance be attained and everyone involved can prosper. To achieve that, it is better if life scientists invite participation from bioinformaticians during the planning phase of their projects.
Finally, measuring the quality of the collaborative relationship is very important. Efforts should be made to find and apply suitable methods to regularly assess such relationships.