This post is an addendum of my previous post where I talked about "purists". According to them, computers are nothing more than a tool to visualize the data they have generated and write the papers. Computational predictions and methods for correct data usage are folkloric characters. Fortunately, there are no more place for "purists" and funding bodies, journal editors, etc. are, from a long time, understanding not only the importance of bioinformatics, biostatistics, etc; also are understanding that the future of science is interdisciplinary. Now to the post.
Not much longer ago, experimentalists had to send material back and forth until they had the necessary data for an analysis. Imagine this scenario not in research, but in a clinic where every second matters to a patients' lives. This is completely different now. Not only computational biology is constantly developing new tools to automate search for [name anything]; but they are also building massive software tools that encompass multiple tools in a user-friendly way, for researchers that are not so familiar with the usual command-line utilities and programming. Furthermore, all of these are using every time more network solutions such as cloud computing; maximizing the usage of resources and facilitating multiple fields of research or applied medicine.
An example of such a massive tool that is user-friendly and encompasses many other smaller tools, in which the user can build his own pipeline without the need for complex programming skills is EDGE (Empowering the Development of Genomics Expertise). This tool is mentioned, in a discussion on "how bioinformatics is bringing genetic analyses to the masses" in a recent article in Nature .
Of course, such massive frameworks are aim of a number of concerns, some of them who really divides researchers on how the best approach for such systematized tool should be implemented and distributed.
The first issue is whether the fundings of such a project drown up. This is an easy one, at least for me. Simply making it open source solves such an issue. What about maintenance of the software? You would be amazed of the power that a community has.
The second issue is the one that bothers me the most. Having a "set-up" "plug-and-play" pipeline for users without much knowledge of programming or command-line utilities definitely, at some time, will lead to "opinionated software". Or in the wise words of Prof. Kathleen Fisch, interim director of the Center for Computational Biology and Bioinformatics at the University of California San Diego:
"Just because you can run the tools doesn’t mean that you should run the tools."
This is where the advice of trained computational biologists come into play. They can give experienced advice, maintain (and create, in the first place) the tools and they are the ones who can really translate virtually any result output by automated softwares. For these reasons, many administrative research bodies are understanding that computational biology is not at it's end. It is at its beginning.
However, the development of such multiscale computational frameworks with a huge infrastructure needs to be a standardized endeavor. As an example I would cite the COST (European Cooperation in Science and Technology) . To facilitate the process of high infrastructure computing, the COST Action CA15120 Open Multiscale Systems Medicine (OpenMultiMed) has been initiated (there were some controversies on that, but i will let that for another post).
The two elements of CA15120 represent the main specific science & technology challenge:
"1. To develop novel multiscale systems medicine concepts, methods and technologies that provide effective, efficient and economical solutions for emerging and future approaches to multiscale systems medicine.
2. To develop a transdisciplinary multiscale systems medicine framework that integrates systems medicine, multiscale modelling, multiscale data science and multiscale computing at the level of research, education and training."
I'll stop for today and re-start the discussion in a proximal future.
1. Perkel J.M. (2017) Nature, 543:137-138.
2. Zanin M. et al. (2017) Briefings in Bioinformatics, bbx160.