Finding. An automated FFPE wash finds more melanoma proteins, more quietly, than a human at the bench, and it does not scramble the proteome. In 54 primary cutaneous melanomas, mostly early stage, tumors that show histological regression look mitochondrially and translationally busier and innate-immune/complement quieter. Survival does not move in this cohort. The tissue still looks like tissue under digital pathology.
Why this paper matters
FFPE archives are the only way most melanoma history exists as protein. Deparaffinization is the step that usually wrecks yield. Automating it without breaking morphology or global profiles is infrastructure. The regression signature is the biology: mitochondria and ribosomes up, complement down.
What they actually measured
54 patients, automated versus manual protein IDs and variance, overlap statistics, regression DE (97/226), pathway enrichment, survival, digital pathology.
How to read the score
Mid sixties. Methods plus a mitochondrial pathway hint. Score 66.
What to do with it
If you run FFPE proteomics, take the Fluent workflow numbers. If you study melanoma regression, pull the mitochondrial and complement protein lists. Do not call regression a survival biomarker from this brief. The directional implication is that regressing early melanomas are proteomically more mitochondrial and less complement-inflamed.