Differential-accessibility scripts

← Previous: Consensus peaks and counting · Workflow contents · Next: Exploratory analysis →

04_run_primary_deseq2.R reproduces the fixed primary analysis from a clean count matrix, peak annotation, and ordered metadata. It uses the 10-in-5 filter, design ~ donor + cell_type, Mature NK reference, Memory-relative-to-Mature contrast, adjusted-p-value threshold 0.05, absolute-LFC threshold 1, and apeglm shrinkage.

The script can install missing R packages into the active user library. Review that behavior and the recorded versions before running it in a new environment. It writes to a new public_reproduction_run directory and refuses to overwrite an existing run.

05_summarize_nominal_sensitivity.R and 06_plot_nominal_sensitivity_volcano.R require the full saved result tables for the primary 10-in-5, 10-in-3, and no-explicit-prefilter analyses. Those large tables are not included in this source-oriented repository. The scripts summarize existing tables; highlighted raw-p-value regions are nominal, not FDR-significant.

---

← Previous: Consensus peaks and counting · Workflow contents · Next: Exploratory analysis →