Fig. 10: Analysis of noise in gene expression of Yeast cells.
From: Advanced methods for gene network identification and noise decomposition from single-cell data
A Distribution of inferred parameters across the cell population. The plots indicate significant variability in system parameters within the yeast population. B Stationary mean and variance of mRNA counts in different cells. We contrast the mean and variance estimates from the inferred models with those estimated from experimental data. The stationary distribution of each real cell was estimated using the occupation time distribution of mRNA measurements. The results show that our estimates are consistent with the experimental data, with all the dots distributed around the diagonal lines. Also, the results demonstrate that the yeast cells displayed considerable heterogeneity in stationary mean and variance. (C) Noise decomposition in stationary distributions. The variance of mRNA counts across the population can be expressed as \({{{{{{{\rm{Var}}}}}}}}\left({X}_{{{{{{{{\rm{mRNA}}}}}}}}}\right)={\mathbb{E}}\left[{{{{{{{\rm{Var}}}}}}}}\left({X}_{{{{{{{{\rm{mRNA}}}}}}}}}| \theta \right)\right]+{{{{{{{\rm{Var}}}}}}}}\left({\mathbb{E}}\left[{X}_{{{{{{{{\rm{mRNA}}}}}}}}}| \theta \right]\right)\), where \({\mathbb{E}}\) and Var are the notations of mean and variance under the stationary distribution, XmRNA is the mRNA count of a randomly selected cell, and θ represents the model parameters of that cell. \({\mathbb{E}}\left[{{{{{{{\rm{Var}}}}}}}}\left({X}_{{{{{{{{\rm{mRNA}}}}}}}}}| \theta \right)\right]\) and \({{{{{{{\rm{Var}}}}}}}}\left({\mathbb{E}}\left[{X}_{{{{{{{{\rm{mRNA}}}}}}}}}| \theta \right]\right)\) are the intrinsic noise and extrinsic noise. We compare these types of noise obtained from the experimental data and inferred models. The plot shows consistency between the noise decomposition results derived in both ways with relative discrepancies about 10%. These discrepancies can be attributed to the limited duration of the measurement period (bringing in bias in the estimates), the exclusion of certain biological mechanisms (e.g., cell cycles), etc. Source data are provided as a Source Data file.
