Biology Reference
In-Depth Information
to the change in perspective associated with this move. A network decomposed
in one of these ways is considered not in a functional, but in a connectionist per-
spective, as known from cognitive science (neural networks composed of large
numbers of units together with weights that measure the strength of connections
between the units but without particular functions ascribed to the units; see, e.g.,
Rumelhart & McClelland, 1986; McClelland & Rumelhart, 1986). Functionality
may be brought in again in 'unbiased' top-down systems biology if functions
are ascribed to the modules in a later modeling step. This, however, is nei-
ther required nor does it change the 'unbiased' picture of a network once it is
obtained.
Abstaining from function ascriptions might be regarded as unusual in biology
(see, e.g., Rosenberg, 1985 on the ubiquity of functional explanation at the level
of macromolecules). We want to point out an important consequence of this
move: Abandoning the functional perspective on biological systems amounts
to 'physicalizing' them, as reflected in the - welcome - improvement of the
mathematical tractability of a system that is decomposed according to explicit
formal criteria. On the other hand, such physicalization renders problematic
the application of a whole range of concepts that are related to the concept of
functionality and that are used, within the realm of natural sciences, exclusively
in biology. In particular, this move eliminates the only reference point for what
might be regarded as the biological 'meaning' of a structure within an organism,
viz. its function within the system or disposition to fulfill a functional role.
'Meaning' has to be put in quotes here anyway, but we cannot think of a weaker
adequate interpretation than that as a systemic function. 16 An approach that
refrains from function ascriptions should not attempt, then, to find anything like
an account of biological 'meaning' in its models (unless having explicated what
'meaning' could be instead). Yet finding biological 'meaning' in 'omic' data is
regarded as one of the main challenges in systems biological modeling (Huang,
2004; Joyce & Palsson, 2006).
16 An evolutionary perspective might allow a different interpretation of the concept of biological meaning, but
none of the systems-biological models we have considered in this chapter make claims to having evolutionary
import. Although at present still somewhat marginal within the field, work that profitably combines a systems-
biological perspective with evolutionary considerations is actually undertaken. Thus, e.g., Huang (2004)
suggests that as “structures midway between genome and phenome”, molecular network topology opens a
new window to study evolution in complex living systems because it sheds new light on the old debate on the
relative contributions of natural selection and self-organization ('intrinsic constraints'), respectively. Another
example is Eric Davidson and his co-workers' research at the interface of systems biology and evolutionary
developmental biology ('EvoDevo'): their research on the control of the development of the animal body
plan by large gene regulatory networks (GRNs) documents how the evolution of body plans must depend
upon change in the hierarchical architecture of developmental GRNs and suggests that the conservation of
phyletic body plans may have been due to the retention since pre-Cambrian time of 'GRN kernels' underlying
development of major body parts (Davidson et al., 2002; Davidson & Erwin, 2006; cf. Callebaut et al., in
press, for a critical assessment of the GRN approach in the larger framework of EvoDevo).
Search WWH ::




Custom Search