Information processing in biology

To survive, organisms must respond appropriately to a variety of
challenges posed by a dynamic and uncertain environment. The
mechanisms underlying such responses can in general be framed as
input-output devices which map environment states (inputs) to
associated responses (output. In this light, it is appealing to
attempt to model these systems using information theory, a well
developed mathematical framework to describe input-output systems.
Under the information theoretical perspective, an organism’s
behavior is fully characterized by the repertoire of its outputs
under different environmental conditions. Due to natural selection,
it is reasonable to assume this input-output mapping has been fine
tuned in such a way as to maximize the organism’s fitness. If that
is the case, it should be possible to abstract away the mechanistic
implementation details and obtain the general principles that lead
to fitness under a certain environment. These can then be used
inferentially to both generate hypotheses about the underlying
implementation as well as predict novel responses under external
perturbations. In this work I use information theory to address the
question of how biological systems generate complex outputs using
relatively simple mechanisms in a robust manner. In particular, I
will examine how communication and distributed processing can lead
to emergent phenomena which allow collective systems to respond in
a much richer way than a single organism could.