Many proteins in living cells are nanomachines that undergo mechanical transformations. Modern modeling methods permit the manipulation of such proteins to discover the physical mechanism behind their function. Applying forces, one can induce geometrical changes characteristic of the proteins' role in the cell and, beyond obtaining qualitative insight, calculate the work W done during a machine cycle. Unfortunately, the manipulations realized in computer modeling include irreversible work which needs to be discounted for comparison with the energetics of naturally occurring, i.e., slower, machine cycles. Recent reports [1 , 2] suggest how this can be achieved mathematically by means of statistical physics. The new methodology is demonstrated on the most simple mechanical transformation of a protein, the stretching of a so-called alpha-helix which behaves like a coiled spring.