A contamination-adjusted tissue thrombopoietic framework identifies a clinically relevant CAF-myeloid niche in non-small cell lung cancer: a paired transcriptomic and survival study

Platelet-related cancer biology is often studied through circulating markers, but tissue-centered thrombopoietic programs in lung cancer are difficult to distinguish from blood-contamination artifacts. In this study, paired bulk RNA-seq profiles were generated from human non-small cell lung cancer tumor tissues and matched adjacent non-tumor lung tissues. Sample identifiers beginning with C denote tumor tissue libraries, and identifiers beginning with N denote matched adjacent non-tumor tissue libraries; shared numeric suffixes indicate patient pairing. These data were used as the local discovery layer for a contamination-adjusted megakaryopoiesis-platelet tumor axis (MPTA) framework. The analysis combined paired differential expression, explicit contamination-aware score construction, public single-cell and spatial reference analyses, external cohort assessment, and local clinicopathologic and survival evaluation. The deposited sequencing data support the tumor-adjacent transcriptomic comparisons and locked tissue-centered MPTA score construction described in the manuscript.