Login to Your Account

Computer science vs. biology at heart of Human Brain Project division


By Nuala Moran
Staff Writer

LONDON – Members of the €1.2 billion (US$1.63 billion) Human Brain Project (HBP) have hit out at criticisms of the project's scientific objectives and governance made by prominent neuroscientists, saying the protest "divides rather than unifies our efforts to understand the brain."

The complaints were made to the European Commission in an open letter which had 156 signatories when it was published on July 7. By Friday, a total of 562 scientists had put their names to the document, which along with threatening a boycott of HBP research, said the project's goals are "overly narrow" and "so unrealistic that they will damage the whole of neuroscience."

Over the course of the week, the head of HBP, Henry Markram of the Ecole Polytechnique Federale de Lausanne, Switzerland, mounted the sole defense of the project. However, on Thursday evening, the members of the HBP, which currently involves 112 research institutions across Europe, united behind him in delivering a riposte.

At a basic level the row boils down to computer scientists vs. biologists. The biologists say not enough is known about the fundamentals of how the brain works to achieve the overarching ambition of the project, which is to create a complete model of the brain in silico.

The computer scientists, on the other hand, argue that the model will provide a technology platform and collaborative research tool around which to unite and integrate the findings of smaller-scale, disparate biology projects. That us-vs.-them battle of the disciplines has destabilized personal relationships, leading neuroscientists to add poor governance and lax supervision of HBP to their list of complaints.

After four years in gestation and following a competition against other proposed mega-projects, the HBP secured its first funding of €54 million last October, for the initial ramp-up phase of the program. It is then due to receive €100 million per annum over 10 years, with half coming from the European Union's €70 billion research program Horizon 2020 to fund the core brain simulation project, and half from national funding agencies for "HBP Partnering Projects."

It was the submission for the second round of European Commission funding in June – in which cognitive neuroscience was moved from the core project to become an HBP Partnering Project – that prompted dissenting neuroscientists to go public.

The June submission "reflected an even further narrowing of goals and funding allocation, including the removal of an entire neuroscience sub-project," the protest letter stated. Unless the European Commission takes action to correct the science and the management of HBP, the neuroscientists said they will boycott the project by refusing to apply for national grants for partnering projects.

In response to the criticism that the goal of a whole-brain simulation is unrealistic because not enough is known to take on such a challenge, members of HBP said that while they share this uncertainty, no one really knows how much neuroscience data are available currently, because those data have never been properly organized. It is also the case that other areas of science have demonstrated that simulation can be a tool to create new knowledge, not just to confirm existing results.

Currently, neuroscience research receives €1 billion in funding per year in Europe and a further $7 billion in the rest of the world. Whilst that is generating lots of new data, there is no technology for sharing, organizing, analyzing or integrating that information. "The HBP will provide the critical missing layer," the project members claimed.

In aiming to do that, HBP is different in kind from the U.S. BRAIN initiative, which is spending $500 million per annum funding independent investigators in a number of fields. "HBP is a data integration project, not a data generation project," the HBP members said. A common platform for data integration is the key missing ingredient in neuroscience.

It is not surprising that computer science is the dominant discipline in HBP, since it has its roots in the Blue Brain project, a collaboration between Markram and computer company IBM, which began in 2005.

The emphasis on computing also is understandable given that HBP sits under the European Union's Future and Emerging Technologies program, which funds frontier research it is hoped will underpin next-generation information and communications technology. While HBP is intended to develop a technology platform for brain research, it is also intended that greater understanding of how the brain functions will inform the design of more efficient computers.

Reading the charges of secrecy, lack of transparency and favoritism that are made by the protesting neuroscientists, it appears that the academic-industrial partnership between Markram and IBM may not have been the most appropriate starting point in terms of governance structures, for such a large-scale, cross-disciplinary pan-European project.

The dissenters made their protest in advance of a review of the project by the European Commission, due to be carried out in January 2015. That is intended to pave the way for the operational phase of HBP to start in 2016. The European Commission did not comment on the content of the letter, but said it will stick to the timetable for the review.