微信号

欧洲一个科学评审小组建议，欧洲人类脑计划（HBP）—— 一项存有争议的“巨无霸”研究项目，必须进行改革以保持正确方向，而且须尽快改革。欧盟3月6日公布的该委员会的一份总结报告称，HBP的管理系统、合作方式以及交流方式等需要进行一系列的“纠正措施”。

这项报告并未直接提出去年欧盟一批神经科学家对HBP项目的反对，但是却提及了他们担忧的一些问题。“我们非常高兴，因为我们提出的问题得到了确认。”批评人士之一、瑞士日内瓦大学计算神经科学家Alexandre Pouget说，“他们正在指出我们曾经指出的问题。”

但是HBP理事会主席Philippe Gillet则表示，这些评审并非都是批评性的，HBP对相关建议表示欢迎。“我们的计划被认为是个好项目，但（评审委员会）给我们提出了一些任务。”他说，“在科学领域，永远不会有完美的评审。”

委员会表示，HBP的管理必须改革，以“确保决策过程简单、公平、透明”，而且HBP旗下的各个项目必须更协调地工作。根据这项5页的报告摘要，HBP必须在2015年6月前作出重要改革，报告还5次提及“尽快”或“从速”等字眼。

HBP是瑞士联邦理工大学神经科学家Henry Markram构思出的两项所谓的“旗舰计划”之一，该计划在欧盟和欧盟成员国之间得到了10亿欧元的经费。然而，HBP的管理系统和优先项目却饱受批评，一些科学家认为，该计划关于绘制整个人类大脑计算机模型的终极目标不切实际。去年，在一份公开发表的信中，数百名神经科学家威胁抵制这项计划，除非欧盟对其彻底审查。

尽管评审委员会希望改善HBP的管理系统，但却没有给出具体的意见；而是希望由一个独立的中间委员会提出这些意见，并期待该委员会在未来几周公开具体意见。然而，评审委员会却表示，HBP必须“更好地描述其战略目标”以及“避免不切实际的想法带来的开销”。

《科学》相关报道

The Human Brain Project (HBP), a humongous, controversial research project backed by the European Union, must reform to stay on course, a review panel has recommended—and it must do so fast. A summary of the panel's report, published today by the European Commission, says a series of "corrective actions" needs to be taken in HBP's governance, the way it collaborates, and its communication.

The report doesn't directly address last year's revolt against HBP by a group of European neuroscientists, but appears to address several of their concerns. "We are very pleased, because it's confirming the problems that we have been pointing out," says computational neuroscientist Alexandre Pouget of the University of Geneva in Switzerland, one of the critics. "They are making the exact same points we have made."

But Philippe Gillet, chairman of the HBP's board of directors, says the review isn't unusually critical and that HBP welcomes the suggestions. "We are ranked as a good project, but [the review panel] gives us some tasks," Gillet says. "You never get a perfect review in science."

The panel says HBP's governance must be changed "to ensure that decision making processes are simple, fair and transparent," and that various HBP subprojects must work together better. Key changes must be implemented by June 2015, according to the 5-page summary, which uses the words "as quickly as possible" or "as soon as possible" five times.

HBP, the brainchild of neuroscientist Henry Markram of the Swiss Federal Institute of Technology in Lausanne, is one of two so-called flagship projects, chosen to receive up to €1 billion in funding from the European Commission and E.U. member states. HBP's management and priorities have come under fire, however, and some scientists have called its ultimate ambition to model the entire brain in computers unrealistic. In an open letter published last year, hundreds of neuroscientists threatened to boycott the project unless the European Commission stepped in to demand an overhaul.

Although the review panel wants to improve HBP's governance, it doesn't give concrete advice; that is expected to come from an independent mediation committee, set up after the rebellion and chaired by Wolfgang Marquardt of the Forschungszentrum Jülich, a research center in Germany. That group is expected to release its own report on HBP within a few weeks.

The review panel does say, however, that HBP must "better articulate its strategic goals" and "avoid at all costs creating unrealistic expectations.” That language is a clear nod to the project’s critics, Pouget says. "The reviewers have said: 'Stop saying unrealistic things; get real.' We're very happy with this," he says.

The panel also says the collaboration between the 13 subprojects must improve. Six of those focus on developing computing platforms and models, whereas four others will amass the experimental data and theories needed to feed them. But so far, "the HBP is not developing with the expected level of integration and the project controls in place are not adequate to achieve this aim," says the summary.

A source within the European Commission says that under E.U. rules, the commission can't release the full review or the names of the scientists who produced it. However, given the controversy and the international interest, the commission wanted at least a summary to be published, says the source, who adds that HBP agreed to this.

Gillet says that the review is "helpful," and that HBP is already making changes to address the panel's concerns. ("We're only 1 year old," he says.) On Tuesday, the project annouced that its three-member executive committee, which included Markram, has been dissolved; from now on, the 22-member board of directors as a whole will make the important decisions. Eventually, HBP wants to evolve into an international organization, Gillet says, modeled on CERN, the European Space Agency, or the European Molecular Biology Laboratory. Three working groups will make recommendations on how to make that transformation, he adds.