may need to be changed as the subject develops. In particular, the GMAG 
should assess any new methods of physical or biological containment that may 
be developed by laboratories and consider whether they would justify major 
modifications of practice. Such assessment would need to be on the basis not 
only of submitted documents but also of independent technical evaluation and 
validation and all interested laboratories should be kept informed of such 
developments. A procedure for the acceptance by the GMAG of new methods 
of containment and of disabled strains might be introduced. The GMAG 
should publish an annual report and be ready to advise on general matters 
connected with the safety of genetic manipulation, including health monitoring, 
for which it should be able to call upon the experience of epidemiologists, and 
the training of staff. 
5.5 At first sight, there would seem to be a case for assigning to the existing 
DPAG the functions which we envisage for the GMAG. But those functions 
go considerably beyond the current activities of DPAG and the factors to be 
considered in assessing genetic manipulation experiments are different from 
those where dangerous pathogens are involved. The control of dangerous 
pathogens involves the application of well-known precautions against a small 
number of easily identifiable and well characterised agents. With genetic 
manipulation on the other hand the hazards and the precautions to be taken 
will depend on the detail of the experiment being undertaken. It will therefore 
be necessary for the GMAG to scrutinise indi\idual scientific proposals in 
detail and perhaps to establish a dialogue with the scientists concerned about, 
for example, particular details of the conditions in the laboratory concerned. 
It seems to us that the roles and methods of working of the proposed GMAG 
and of the existing DPAG will be so different that the combination of the two 
roles as the responsibility of a single group would reduce the efficiency with 
which both tasks were carried out. We therefore recommend that the GMAG 
should be separate from the DPAG but that, because comparable laboratory 
containment measures are key safety features in both fields, there should be 
liaison between the two groups: in particular, if a proposed genetic manipula- 
tion experiment involved the use of a dangerous pathogen in the DPAG’s 
Category A, then DPAG procedures for control would apply, although the 
GMAG would clearly need to be involved as well. 
5.6 We envisage the following stages in the consideration of a proposal for 
an experiment involving genetic manipulation: 
i. discussions within the laboratory both of scientific merits and of 
potential hazards. The laboratory's Biological Safety Officer and a 
properly constituted and representative safety committee should have 
key roles to play and the discussions should lead to provisional 
conclusions about the desirability of conducting the experiment and 
about the containment category into which the experiment should 
fall; 
ii. if as a result of these discussions there is no doubt that the proposed 
experiment belongs to Category I or II the GMAG should be notified 
immediately, but work could proceed under the appropriate conditions 
as specified in the code of practice; 
iii. if the discussions suggest that the proposed experiment falls into 
category III or IV, reference must be made to the GMAG for advice 
before a final decision to undertake the experiment is taken; 
IV-16 
[ 456 ] 
