Code and derived data for "AGR2 tracks mucin-client demand across adenocarcinomas, revealing lineage-tuned proteostasis"

This record contains the derived data and analysis code supporting the manuscript “AGR2 tracks mucin-client demand across adenocarcinomas, revealing lineage-tuned proteostasis.” The study investigates whether deployment of the endoplasmic-reticulum protein disulfide isomerase AGR2 is quantitatively coupled to lineage-specific secretory client demand rather than simply to generic ER folding and redox machinery across epithelial cancers. The archive contains: the locked per-sample score table used for manuscript-level analyses (Supplementary Data S1); derived statistical and cohort-level analysis tables; gene-set and scoring definitions; Python analysis scripts and archived analysis notebooks; scripts used to reconstruct figures and supplementary tables; reliability-matched null analyses and dependent-correlation tests; lineage- and molecular-subtype analyses; EMT and proteostasis-module analyses; GDF15-related analyses; patient-level single-nucleus and spatial analyses; documentation of dataset provenance, local input mapping and computational dependencies. The submitted analyses use the locked 22-gene AGR3-free SecretoryFactor definition. Primary third-party expression, proteomic, single-cell and spatial datasets are not redistributed in this archive. Their originating public repositories, accession identifiers and local analysis-input mappings are documented in docs/DATA_SOURCES.md and docs/INPUT_FILE_MAP.tsv. Figure-level Source Data are submitted separately with the manuscript and are not duplicated in this archive