inmoose.deseq2.estimateDispersionsMAP

inmoose.deseq2.estimateDispersionsMAP(obj, outlierSD=2, dispPriorVar=None, minDisp=1e-08, kappa_0=1, dispTol=1e-06, maxit=100, useCR=True, weightThreshold=0.01, modelMatrix=None, type_='DESeq2', quiet=False)

Low-level function to fit dispersion estimates

Normal users should instead use DESeqDataSet.estimateDispersions(). This low-level function is called by DESeqDataSet.estimateDispersions(), but is exported and documented for non-standard usage. For instance, it is possible to replace fitted values with a custom fit and continue with the maximum a posteriori dispersion estimate.

estimateDispersionsPriorVar() is called inside this function, and stores the dispersion prior variance as an attribute of DESeqDataSet.dispersionFunction, which can be manually provided to estimateDispersionsMAP for parallel execution.

Parameters:

  • obj (DESeqDataSet) – the input dataset

  • outlierSD (int) – the number of standard deviations of log gene-wise estimates above the prior mean (fitted value), above which dispersion estimates will be labelled outliers. Outliers will keep their original value and not be shrunk using the prior.

  • dispPriorVar (array-like) – the variance of the normal priori on the log dispersions. If not supplied, it is calculated as the difference between the mean squared residuals of gene-wise estimates to the fitted dispersion and the expected sampling variance of the log dispersion.

  • minDisp (float) – small value for the minimum dispersion, to allow for calculations in log scale, one order of magnitude above this value is used as a test for inclusion in mean-dispersion fitting

  • kappa_0 (float) – parameter used in setting the initial proposal in backtracking search, higher kappa_0 results in larger steps

  • dispTol (float) – parameter to test for convergence of log dispersion, stop when increase in log posterior is less than dispTol

  • maxit (int) – maximum number of iterations to allow for convergence

  • useCR (bool) – whether to use Cox-Reid adjustment

  • weightThreshold (float) – threshold for subsetting the design matrix and GLM weights for calculating the Cox-Reid correction

  • quiet (bool) – whether to print messages at each step

  • modelMatrix (array-like) – for advanced use only, a substitute model matrix for gene-wise and MAP dispersion estimation

  • type ("DESeq2" or "glmGamPoi") – specify if the glmGamPoi package is used to calculate the dispersion. This can be significantly faster if there are many replicates with small counts.

Returns:

the input obj with final MAP dispersion estimates

Return type:

DESeqDataSet