rna_majiq.PsiOutliers

class rna_majiq.PsiOutliers(cases, controls, alpha_case=None)

Outliers in PSI between cases and controls

Parameters:

cases (Union[PsiCoverage, PsiGroup]) – PsiCoverage for prefixes that will be treated as cases
controls (PsiControlsSummary) – Summary of PsiCoverage for prefixes that are treated as controls
alpha_case (Union[float, None]) – Optional single float between 0 and 1 to use for cases instead of the controls_alpha array

Methods

`__init__`(cases, controls[, alpha_case])
`abs_dpsi_mask`(threshold)	boolean array indicating if event connection abs_dpsi passes threshold
`summarize_df_events`(df_ecidx[, ...])	Summarize table in format of `PsiOutliers.to_dataframe()` to events
`summarize_df_genes`(df_events[, ...])	Summarize table from `PsiOutliers.summarize_df_events()` to genes
`to_dataframe`([sg, controls_min_experiments, ...])	Return table of comparisons between cases and controls

Attributes

`abs_dpsi`	abs(dPSI) from case posterior mean vs controls population median.
`cases_psi_quantile`	Posterior quantiles of PSI (approximate posterior)
`cases_q`	Which quantiles will be evaluated for cases
`dpsi_lb`	dPSI from case quantiles vs controls median.
`dpsi_lb_scaled`	dPSI of case quantiles vs controls median scaled by controls interquantile range.
`dpsi_quantile_gap`	Gap in PSI between extreme quantiles of cases vs controls.
`tail_probability`	Probability that cases psi < ub quantile or psi > lb quantile