inmoose.deseq2.makeExampleDESeqDataSet

inmoose.deseq2.makeExampleDESeqDataSet(n=1000, m=12, betaSD=0, interceptMean=4, interceptSD=2, dispMeanRel=<function <lambda>>, sizeFactors=None, seed=None)

Make a simulated DESeqDataSet

This function constructs a simulated dataset of Negative Binomial data from two conditions. By default, there are no fold changes between the two conditions, but this can be adjusted with the betaSD argument.

See also

sim_rnaseq

another function to create simulated Negative Binomial data, accounting for conditions, outliers and batches

Parameters:

  • n (int) – the number of genes

  • m (int) – the number of samples

  • betaSD (float) – the standard deviation for non-intercept betas, i.e. \(\\beta \sim \mathcal{N}(0, betaSD)\)

  • interceptMean (float) – the mean of the intercept betas (log2 scale)

  • interceptSD (float) – the standard deviation of the intercept betas (log2 scale)

  • dispMeanRel – a function specifying the relationship of the dispersions on the 2^trueIntercept

  • sizeFactors (array-like) – multiplicative factors for each sample

Returns:

a DESeqDataSet with true dispersion, intercept and beta values in DESeqDataSet.var. Note that the true betas are provided on the log2 scale.

Return type:

DESeqDataSet