The Secret of Building a Scientific Community

Scientific communities are the base on which most scientific projects are built, particularly in the life sciences. Scientific networks (Fig. 1) are required because all projects are complex and the skills required for their successful accomplishment are more than that of what a single individual can contribute.

A sketch of a small network displaying community structure, with three groups of nodes with dense internal connections and sparser connections between groups. [Taken from Wikipedia; author j_ham3 (CC BY-SA 3.0)]
Fig. 1: A sketch of a small network displaying community structure, with three groups of people with dense internal connections and sparser connections between groups. [Taken from Wikipedia; author j_ham3 (CC BY-SA 3.0)]
I believe the success of current and future scientific careers is going to be conditioned more and more by the ability of managing relations in these communities, be able to influence them and understand their dynamics. In fact, since I started doing proper research 14 years ago, I do not remember a single research project I have been involved in which I was not part of a scientific network. This trend has only increased in the past decade by the embrace of the myriad of communication channels now available (chat, Skype, hangouts) together with more traditional ones (email, phone, post), making it possible for networks of scientists to collaborate on line.

Scientific networks have fascinated me from very early on because I saw a great potential in them. I started the International Society for Computational Biology Student Council, the leading student organisation in the field. After that I participated in other networks, but the one now really exciting me is the BioJavaScript (BioJS) open source community, in which I currently have the privilege to be its coordinator. Currently BioJS has around 20 developers who have contributed code, about 120 people in the BioJS google group list and our first paper published in 2013 has 15 citations to date. We have our code freely available in GitHub and its main page there receives about 1000 visits and 32 code downloads per month (Figure 2). I believe these metrics are quite substantial, particularly if we take into account that this community has been formally established for about 2 years only.

GitHub portal for BioJS. It shows 621 commits, 9 branches, 2 releases and 20 contributors.
Fig. 2: GitHub portal for BioJS. It shows 621 commits, 9 branches, 2 releases and 20 contributors.

Sometimes I am asked about how the BioJS community has been built and what is the secret of its success. This question is often accompanied by more specific ones such as how I handle the exposure of the community or how I keep its members motivated.

I do not claim to have all the answers to these questions, but in this post I would like to summarise some of the things I have learned through my experience in building this community. Key success factors in managing the BioJS community are based on the following principles:

1. Publish often and be generous with authorships. I think our community papers have been by far the greatest community building exercise that we have carried out. People feel really rewarded even though we do not have any formal funding (or hardly any). Including people’s names in papers (as long as they have minimally contributed to the code and participated of our discussions) has been a great strategical decision.

2. Keep the communication flowing. As a matter of principle the core 20 people of the project talk once a month. We always meet the first Thursday of the month at a time that is compatible with European and Californian time zones. This is also a great exercise in community building. I always make sure that we have an agenda with the achievements and todo things for the time being. Another important thing is that anyone attending these calls can bring anybody they think that could contribute to the call. Although I coordinate these calls, I do not decide who can attend, we keep participation to this call pretty open.

3. Delegate responsibilities early. I have been quite fortunate to be surrounded by very competent people from the beginning who have worked really hard. As the project has grown, it was clear that not everything could be discussed in the monthly calls, so we decided to create committees. These committees would specialise on special areas of interest such as technical, production, outreach/training and Google Summer of Code internships (GSoC; Fig. 3). The result is that committee leaders step up, take responsibility and begin to own the project. This is crucial. As a coordinator or leader of an open community you may feel that you lose control or power by allowing people to become the champions of certain areas. In fact, by giving people the chance to lead, I have found they contribute more, are more motivated and they become loyal to the project. Although I cannot claim that I am the person who knows most about the technical aspects of the project, I know that the technical or production or outreach or GSoC leaders will listen to me. This is one of the most rewarding feelings about being a coordinator for this open community.

People leading the Google Summer of Code team for BioJS.
Fig. 3: People leading the Google Summer of Code team for BioJS.

4. Pay close attention to every single opportunity to raise awareness of the project. This may sound very obvious, but my experience is that many developers, scientists or collaborators do not consider publicity, marketing and outreach as something “fun” or a good use of their time. Nothing is further from the truth. I have always found that any time spent on promoting, evangelising and networking is one of the most fruitful investments in the project. What is the trick for really attracting more people and attention to the project? If you are really excited and believe in the project, that is the greatest way to attract people to the project. It really pays off to be in the look out for new places to present your project, including conference presentations, blog entries, tweets, tutorials, hackathons, etc.

5. Be in the constant look out for new influential research groups or companies who could join your community. You cannot take over the world all at once. You need to target your efforts to your next level in your circle of influence. Something that works really well for me is to engage leading scientists through article reviews, requests to become scientific advisors, opportunities to be co-PI in grants, etc. But all of the above points cannot really be effective unless you…

6. Have fun. Really. Unless you enjoy the project itself and believe in it you will never make it into a real community.

To summarise, what’s the secret really? Here is the answer: surround yourself by the right people. And the people you need to be surrounded by need to be some combination of the following qualities: passionate, driven, skilful, social and committed. The rest, it simply happens.

1 comment

Leave a Reply