Supplement: Statistical Methodology in Single-Molecule Experiments

The lacking of suitable techniques to manipulate and visualize single molecules was one of the major obstacles to studying microscopic systems. Although high-resolution techniques such as electron microscopes and X-ray diffraction can allow scientists to visualize static molecules bound to hard surfaces or confined in crystallized structures, these tools are not suitable for monitoring the dynamics of single molecules in liquid solutions, especially molecules in biological systems. The earliest modern single-molecule experiments were carried out on ion channels in the 1970s. In these experiments, the patch clamp technique (Neher and Sakmann, 1976) enabled reliable recording of membrane protein conductance in a single ion channel. Afterward, technological advancements drastically expanded scientists’ tool sets for conducting experiments at the single-molecule level. First, new tools were created to manipulate individual molecules. The famous “optical tweezer” enabled scientists to trap or move viruses and bacteria with the force generated by laser beams (Ashkin and Dziedzic, 1987), and could be used to measure the forces and locations of microparticles (Ashkin et al., 1990). Microneedles were utilized to apply and measure force in biomolecules (Kishino and Yanagida, 1988). Force exerted by flow or magnetic field can also serve as a tool to stretch long polymers such as a DNA string (Smith, Finzi, and Bustamante, 1992; Perkins et al., 1995). More importantly, developments in fluorescence microscopy fundamentally revolutionized the visualization of microscopic systems. Fluorescence microscopy utilizes fluorophores that emit photons upon light excitation to illuminate molecules so that their location and structure can be measured using an optical microscope. The application of fluorescence microscopy dates to 1961 (Rotman, 1961). In 1976, single molecules were first imaged at room temperature with the help of dozens of fluorophores (Hirschfeld, 1976). Single-fluorophore level detection was Chao Du is Assistant Professor in Statistics, University of Virginia, Halsey Hall, Charlottesville, VA 22903, USA (email: cd2wb@ virginia. edu ). S. C. Kou is Professor of Statistics and Professor of Biostatistics, Harvard University, Science Center, 1 Oxford Street, Cambridge, MA 02138, USA (email: kou@ stat. harvard. edu ).