Biomolecular machines

What Sustains Life Becomes What Sustains Society (Biomolecular Machines as Advanced Materials for the Future)... 

D.W. Urry, Elastic Biomolecular Machines Synthetic Chains of Amino Acids, Patterned After Those in Connective Tissue, can Transform Heat and Chemical Energy into Motion. Sci. Am. January 1995,64-69. 

Summary. Cells are a collection of machines with a wide range of functions. Most of these machines are proteins. To understand their mechanisms, a synergistic combination of experiments and computer simulations is required. Some underlying concepts concerning proteins involved in such machines and their motions are presented. An essential element is that the conformational changes required for machine function are built into the structure by evolution. Specific biomolecular motors (kinesin and Fi—ATPase) are considered and how they work is described. 

The optimal choice of preconditioner will ultimately depend on the computer architecture, in as much as some are more readily vectorizable or parallelizable. For example, the initial incomplete Cholesky decomposition methods work well on serial machines, but the forward and backward substitutions are not vectorizable. Simpler decompositions, such as diagonal scaling, run faster on machines like the Cray YMP. More complicated, vectorizable variations of the incomplete Cholesky decompositions have been developed (see, e.g., ref. 24) and are currently under investigation for their applicability to problems in biomolecular electrostatics. Studies of multigridding techniques are also very exciting. 

The outline of this chapter is as follows. The second section introduces a number of important supervised learning problems and illustrates how a biomolec-ular application can be cast in each problem formulation. Specifically, modeling protein-DNA interactions serves as the example for each of these formulations. The third section summarizes recent applications of machine learning to biomolec-ular modeling. The final section discusses current trends and future directions of machine learning applications to biomolecular modeling. 

Overall, machine learning is a powerful fool in computer science and will find more and more applicafions in biomolecular modeling. When combined wifh biochemical and biophysical undersfanding, fhis new approach is expected to yield phenomenal advancement in our understanding of protein structures, functions, interactions and localizations in the years to come. 

In the same year, our group in Lausanne published first results from a similar instrument which was equipped with an electrospray ion source for producing closed-shell biomolecular ions, the first demonstrations of which were the measurement of the UV spectra of cold, protmiated aromatic amino acids, tryptophan [46], tyrosine [46, 122], and phenylalanine [122]. Spectroscopic detection is achieved by measuring the small percentage of parent ions that fragment subsequent to UV absorption. The internal temperature of the ions was estimated to be 11-16 K from an analysis of the intensity of hot band transitions of low frequency vibrational modes. If the temperatures achieved in buffer-gas cooled ion traps are low enough and the spectra sufficiently simple, one can often resolve UV absorption spectra for different stable cOTiformers of the molecule [122]. In this case, one can use the IR-UV double resonance techniques so profitably employed in supersonic molecular beam studies [91,123-128] to measure conformer-specific infrared spectra, and this was applied by Steams et al. to both individual amino acids [129] as well as peptides with up to 12 amino acid residues [130]. Subsequent improvements to the Lausanne machine (Fig. 7) included the addition of an ion funnel to... 

---