Perform optional post-hoc filtering and run two-staged statistical tests.
posthoc_and_stager(dturtle, ofdr = 0.05, posthoc = 0.1)
| dturtle | Result object of |
|---|---|
| ofdr | Overall false discovery rate (OFDR) threshold. |
| posthoc | Specify the minimal standard deviation of a transcripts proportion level that should be kept when performing post-hoc filtering. To disable posthoc filtering 0 or |
An extended dturtle object. Additional slots include:
sig_gene: A character vector of all genes where the first stageR step was significant. Basically the significant genes, that showed signs of DTU.
sig_tx : A named character vector of transcripts where the second stageR step was significant. Basically the significant transcripts of the significant genes.
FDR_table : A data frame of the stage-wise adjusted p-values for all genes/transcripts. Might contain NA-values, as transcript level p-values are not available when the gene level test was not significant.
The post-hoc filter excludes transcripts, which standard deviation of the proportion per cell/sample is below the threshold. The two-staged statistical test performed by stageR first determines if any of the transcripts of a gene is showing signs of DTU. The second step tries to identify on singular transcript level the significantly different transcripts.
run_drimseq() for DTU object creation. create_dtu_table() for result table creation.
Other DTUrtle DTU:
combine_to_matrix(),
import_counts(),
priming_bias_detection_probability(),
run_drimseq()