Think: Theory for Africa

1 Division of Cell Biology, Department of Human Biology, Neuroscience Institute and Institute of Infectious Disease and Molecular Medicine, Faculty of Health Sciences, University of Cape Town, Cape Town, South Africa, 2 School of Computer Science and Applied Mathematics, University of the Witwatersrand, Johannesburg, South Africa, 3 Gatsby Computational Neuroscience Unit, University College London, London, United Kingdom, 4 Centre for Neural Circuits and Behaviour, Department of Physiology, Anatomy and Genetics, University of Oxford, Oxford, United Kingdom

perspective, the subspecialty best placed to leverage these developments is computational neuroscience.
The limiting factor for computational neuroscience development in Africa (and anywhere in the world) is quantitative training. Although Africa does not currently contribute a considerable fraction of global neuroscience output (see Fig 1), education in the traditional quantitative disciplines of mathematics, physics, and biology is increasing. African researchers and collaborators are progressively initiating strategies to develop and grow neuroscience expertise through national [9,10], research-domain [11], and education and/or outreach programs [12]. Research-level capacity in quantitative disciplines is being enhanced throughout Africa by initiatives such as the African Institute for Mathematical Science (AIMS), Square Kilometer Array (SKA) [13], the Human Hereditary & Health in Africa (H3Africa) [14], and more neuroscience-focused training organisations such as TReND in Africa and the IBRO. Hallmark features of these initiatives are the integration of local and international expertise as well as sharing of resources. We note also that the confluence of genomics and neuroscience has rich potential for generating unique insights and increasing African capacity in neuroscience [15]. This burgeoning emphasis on quantitative skills development, combined with improved access to data and theory, means that now is the time to put greater effort into facilitating the involvement of Africans in computational neuroscience.
Several promising educational initiatives have recently emerged for Africa. These include annual schools in computational neuroscience (IBRO-Simons Computational Neuroscience Imbizo, isiCNI, almost 100 students trained to date) and machine learning (Deep Learning Indaba, DL Indaba, nearly 1,000 attendees), which both started in 2017. These initiatives bring young, predominantly African scientists together with leading international experts in their field. It is the relationships between students, and particularly between students and faculty, that will support the careers of these fledgling, African computational neuroscientists and machine learning practitioners.
Entering the field of computational neuroscience can be a daunting prospect; there are a proliferation of online courses, numerous textbooks, diverse conferences, and several summer schools. In addition, mailing lists and social media have become an integral component for engaging with current discussions in the field, connecting with authors, and sharing opportunities and information. Although the type of quantitative training background can inform a focus in the field, there are valuable general principles to be gained from the resources in Table 1, especially for those unsure of where to start. Lastly, although research in computational neuroscience is significantly less expensive than experimental neuroscience, there is still some cost involved, mostly for computing resources. Fortunately, many universities in Africa do have high performance computing clusters, and there is the freely available Neuroscience Gateway [16], which offers computational neuroscience-specific computing resources.
Thus far, the major strides in computational neuroscience development in Africa have come from foreign sources. To retain these initiatives and benefit from these pilot projects, African governments and academic institutions must also be willing to support projects, both financially and logistically. In addition to increasing focus on the field for African scientists, we provide several recommendations for the global computational neuroscience community to enhance the field in Africa: 1. Make funding available for Africans to attend international computational neuroscience conferences, workshops, and summer schools by providing cheaper (or free) registration and travel grants.
2. Support or organise computational neuroscience schools and workshops in Africa. In the century of the brain, African scientists and educators are poised to make important contributions to global neuroscience research. We believe that theoretical sciences, and specifically computational and theoretical neuroscience are the ideal discipline for the African continent.