Noise Reduction in Correlation of Fluorescent Time Series Data
College of Arts and Sciences
Fluorescence correlation microscopy stands as a powerful tool for analyzing transcriptional kinetics. However, with modern methods involving increased temporal resolution and decreased fluorescent signals comes the challenge of picking up a statistically significant correlation from the random background in the nucleus of living cells. Using orbital tracking we are able to track a transcription site and its resulting fluorescent signals, however macroscopic changes in the fluorescence intensity within the nucleus can mask the correlation of the signals under examination. While uncorrelated noise should average as zero in an ideal correlation analysis, correlated noise can manifest as extra correlation features. We discuss novel methods using matrix algebra and entropy for the discrimination between correlated noise and signal while applying it to experimental data taken in living cells.
Tisdale, Garrett, "Noise Reduction in Correlation of Fluorescent Time Series Data" (2018). 2018 Undergraduate Research and Scholarship Conference. 81.