Organized systems such as living organisms or cells require that the l
evel of crucial elements somehow be evaluated and taken into account t
o determine their future rate of production. This regulation is effect
ed by feedback loops, which are oriented circuits of interactions. The
re are two types of feedback loops, negative and positive, depending o
n the parity of the number of negative interactions, Their role are ra
dically different: negative loops promote homeostasis while positive l
oops permit multistationarity, with its biological modality, different
iation. Appropriate combinations of positive and negative feedback loo
ps may generate extremely complex behaviour. Real regulatory systems u
sually comprise several feedback loops which may be intertwined in com
plex ways. A major aspect of our work has been the development of meth
ods which permit to derive the dynamics of such complex networks from
their structure. Recent progress has shown that a networks can be trea
ted (without loss of rigor) as the set of its interacting loops rather
than as the set of all its interacting elements, This is somewhat lik
e considering the wheels of a clock rather than the individual teeth w
hich compose the wheels. This approach is illustrated by various concr
ete biological examples.