PARAMETRIC CHARACTERIZATION OF CALCIUM SIGNALS
We have developed a MATLAB-based algorithm in the form of a software that allows for the meta analysis of calcium signals, i.e., the automated characterization of single
and multi-peaked transient calcium responses from time series data. Signals are decomposed into drift and response components and the entirety of the
signal is reconstructed using a mathematical model of a transient response and a drift. The software measures the time of onset, activation time
(10% to 90% of total amplitude), amplitude of response, full width at half maximum and decay time constant of a transient non-oscillatory response.
For oscillatory multipeaked responses, the algorithm attempts to separate coherent oscillations from those that are completely stochastic. For the full
set of oscillations as well as the coherent subset, the algorithm then reports the number of oscillations, period of oscillation (averaged), standard
deviation of the period, magnitude (averaged), full width half maximum (averaged), duty cycle (averaged), and the length of oscillatory persistence.
Time series data should be stored in an excel sheet, and the whole document can be processed with the MATLAB command: characterizeDocument('example.xlsx').
The paper that describes the mathematical theory associated with the algorithm is available:
L. Mackay, N. Mikolajewicz, S.V. Komarova and A. Khadra. Systematic Characterization of Dynamic
Parameters of Intracellular Calcium Signals. Frontier in Physiology, 7, 525, 2016.
Detailed description of the software is available here.