Mutual information between in- and output trajectories of biochemical networks

Biochemical networks can respond to temporal characteristics of time-varyingsignals. To understand how reliably biochemical networks can transmitinformation we must consider how an input signal as a function of time--theinput trajectory--can be mapped onto an output trajectory. Here we estimate themutual information between in- and output trajectories using a Gaussian model.We study how reliably the chemotaxis network of E. coli can transmitinformation on the ligand concentration to the flagellar motor, and find theinput power spectrum that maximizes the information transmission rate.