The noise-decomposition technique is applied in several fields, including genetic systems, optical images, recording, and navigation. In genetic systems, noise decomposition is usually achieved by using two reporters [Elowitz M.B., Levine A.J., Siggia E.D., Swain P·S., 2002. Stochastic gene expression in a single cell. Science 297, 1183–6.]. A reporter is a protein with fluorescence, an RNA hybridized with a fluorescent probe, or any other detectable intracellular component. If a reporter is constructed in addition to the original reporter, the system's stochasticity may change. Such phenomena became severe for genes in plasmids with a high copy number. By SSA (stochastic simulation algorithm), we observed an approximately 50% increment in the coefficient of variation while introducing additional reporters. Besides, if two reporters respond to the upstream element at a different time, the trunk noise (or extrinsic noise) cannot be accurately determined. This is because the “calculative trunk noise” changes along with the delay, though the real trunk noise does not. For RNA reporters, a 5-min transcriptional delay caused a calculative trunk noise that was 90% less than the real trunk noise. Fortunately, this problem is negligible when the degradation rate constant is low, and it is usually true in the case of the protein reporters. One can check the lifespan of the reporter before applying the noise-decomposition technique.

噪声分解技术应用于多个领域，包括遗传系统、光学图像、记录和导航。在遗传系统中，噪声分解通常通过使用两个报告器来实现[Elowitz MB, Levine AJ, Siggia ED, Swain P·S., 2002。单细胞中的随机基因表达。科学 297, 1183–6.]。报告基因是具有荧光的蛋白质、与荧光探针杂交的RNA或任何其他可检测的细胞内成分。如果除了原始报告器之外还构建了报告器，系统的随机性可能会改变。对于高拷贝数质粒中的基因来说，这种现象变得严重。通过 SSA（随机模拟算法），我们在引入额外的报告基因时观察到变异系数增加了大约 50%。此外，如果两个报告器在不同时间响应上游元件，则无法准确确定干线噪声（或外部噪声）。这是因为“计算性干线噪声”会随着延迟而变化，而真正的干线噪声却不会。对于 RNA 记者来说，5 分钟的转录延迟会导致计算性的主干噪声比真实的主干噪声低 90%。幸运的是，当降解速率常数较低时，这个问题可以忽略不计，对于蛋白质报道分子来说通常是这样。在应用噪声分解技术之前，可以检查报告器的寿命。