Sir

A recent News story reported that a US-based “multi-laboratory, multidisciplinary initiative” called the Alliance for Cellular Signalling (AFCS) seeks to provide a more integrated view of cell-signalling pathways ( Nature, 402, 219; 1999). To accomplish this, the AFCS hopes to “map how molecules in a cell interact” without bias towards particular proteins, through the collaborative efforts of systems engineers, biologists and informaticists. This project is to have a budget of approximately US$100 million provided over ten years by the National Institutes of Health (NIH) and private companies.

We would like to express concern regarding both the outline of the experimental systems and the structural organization of the proposed alliance. Despite planning an integrative approach, the organization's website (http://afcs.swmed.edu ) says that “the experimental efforts of the alliance will be focused on two cells that display interesting and important G-protein-regulated phenomena: the B lymphocyte … and the cardiac myocyte”. Although these cell types are of obvious biomedical and pharmaceutical relevance, how can such a biased and limited view contribute to a more global perspective on cell signalling? Such an endeavour is structured without consideration of additional cell types, model systems and other interesting signalling phenomena not necessarily involving G-protein-mediated processes.

We are concerned that the hierarchical structure of this initiative creates an unnecessary division of labour and multiple layers of bureaucracy with too many committees and directors. After a selection process (based on “recognition of valued accomplishments”, according to the website) each member will be required to “contribute detailed, standardized, and quality-controlled information (from the literature) about their assigned molecules”. Besides defining “new molecules of interest”, the governing committees (“whose chairs and members have already been chosen”) will decide on the termination of membership based on an undefined concept of “failure to perform”.

It seems to us that this Orwellian structure minimizes the possibility of innovation and creative approaches towards a real understanding of cellular signalling. Furthermore, a ‘one-member—one protein’ paradigm monopolizes the contributions and marginalizes investigators in the field who are not members.

Members will be required not to publish their results in journals but rather deposit the data on the alliance's website in the form of a “molecule page” bearing their name. Although such an approach may be an efficient means of disseminating information, we feel that eliminating the impartial process of outside peer review is unsound and unacceptable. Moreover, it is unclear to us how the efforts of students and postdocs involved in the actual work will be acknowledged.

Finally, we are most concerned that this “new way of doing business [sic] that requires collaborators to act altruistically”, is restricted to a network of laboratories in North America. As stated, “collaborators cannot be spread across too many time zones”, because the alliance plans to hold teleconferences using Internet 2, which does not yet exist outside the United States. We find this reasoning shallow — and insulting to the international scientific community. The proposed research does not have to be communicated in real-time, and the Internet has worked very well for other large-scale collaborative efforts, notably the sequencing of genomes.

In view of these concerns we question whether the alliance, as it stands, deserves the genuine interest of the scientific community and the requested funding by the NIH, and whether it is such a “new research initiative” after all.