Effect of topology and time window on probability distribution underlying baclofen induced Ca2+ response in hippocampal neurons

Cytosolic Ca2+ oscillation in neurons regulate a wide range of cellular functions starting from cell division to apoptosis. The major challenge in analyzing the large-scale heterogeneous calcium data obtained from hippocampal neurons is that there is no specific tool available for probability density function (pdf) fitting and model ranking. First, we focus on the ranking of various pdf and selection of a particular pdf using maximum log-likelihood. Five pdfs were analyzed in this study, exponential, gamma, log-normal, Rayleigh, and Weibull. Next, we used the statistical models to find the effect of two factors, the network topology and time window for the calcium response. The robustness of the best pdf was validated using multiple datasets obtained through random sampling of neurons from a neuron pool. GPCR targeting drug, baclofen was chosen as the model drug to inhibit Ca2+ response. Strongly-connected neurons show a significant change in Ca2+ oscillations after the addition of drug in comparison to weakly-connected neurons. The proposed technique can be used to study the dose-response from a large number of calcium imaging videos having heterogeneous responses.