Detection and correction of jumps in real-time pcr signals

Abstract

Systems, methods, and apparatuses are provided for detecting and potentially invalidating or correcting jump errors in data from growth processes. A jump error can be identified by determining a second derivative of the data set, and identifying two consecutive cycles with opposite signs in the second derivative. Once a jump error has been detected, the data set can be invalidated or corrected based on various criteria. Whether to invalidate or correct can be based on an absolute jump height, a relative jump height (e.g., relative to the net growth or relative to the baseline), an absolute location (cycle number) of the jump, or a relative location. In one implementation, the jump can be corrected by subtracting a jump height from points subsequent to the jump or by adding the jump height to points prior to the jump.