An automated technique for analysis of current transitions in multilevel single-channel recordings.

Detailed kinetic studies of ion channel gating are best carried out using the patch-clamp technique which permits the measurement of the ionic current through individual channels. Typical patch-clamp recordings show the current signal, in the form of a sequence of rectangular pulses (analogous to a random telegraph signal), riding on slow baseline drift, partially obscured by high-frequency noise and distorted by filtering. In order to analyze such recordings, we have developed a set of interactive Pascal programs based on a feature-detection algorithm capable of identifying current transitions in multiple-channel recordings in the presence of substantial levels of noise and drift. Software operation is largely automated but includes provisions for examination and correction of the output. The software was optimized and systematically evaluated using simulated data with variable amounts of noise and drift. Results indicate that satisfactory performance is obtained for signal-to-noise ratio as low as four even with uncommonly large baseline drift. Steady-state processing speeds varied from 1,000 to 4,000 samples per second depending on data complexity.