Aliquot Dependence of Five Protein Directions in Mitochondrial-Disease Fibroblasts

**Objective:** Determine which simultaneous GPX4-down, MORF4L1-down, MOXD1-down, MSRA-down and TMED9-up abundance contrasts depend on patient-aliquot or control composition in the 24 author-labelled PXD055473 fibroblast cases. The increment beyond marker discovery and source quality control is a casewise decomposition of explicit omission witnesses. **Methods:** An exploratory reanalysis used processed whole-cell protein-group quantities from 72 patient-aliquot runs and 15 control runs representing five recurring controls. Within-batch differences of mean log2 quantities were evaluated under one no-omission, three patient-only, five control-only and 15 joint configurations per case. The available-value point baseline was distinguished from an eligibility-aware configuration baseline requiring initially at least two positive-finite patient values and three controls per marker. Unknown components were retained; any known false component established a nonmatch. All 576 configurations were enumerated without population tests or sampling intervals. **Results:** The available-value point baseline had eight matches among 24 cases. The eligibility-aware baseline had seven MATCH, 16 NONMATCH and one UNKNOWN, because VC7 had ineligible MOXD1. Only VC3 and VC4 remained MATCH throughout. Patient-only omissions supplied known nonmatch witnesses for the other five eligible baseline matches and for VC7. VC8 remained MATCH under all five control-only omissions, whereas omitting patient aliquot 29 changed MORF4L1 from −0.086115 to +0.079545 log2 units. Across all configurations, 312 contained an unknown marker but only 19 had an UNKNOWN conjunction. The subset export reproduced all 106 available deposited marker contrasts; an explicitly imputed export reproduced none, without isolating imputation as the cause. **Conclusions:** A full-run sign match does not establish composition invariance in this deposit. The added tables identify the aliquot-specific qualification needed before casewise research screening; no improved screening performance was tested. Corrected source mapping or reconstruction that removes the witnesses would defeat the result. Processed abundance, finite omission robustness and reproducible arithmetic establish neither a biological mechanism nor diagnostic validity.