# C11 sensitivity analysis: retention of NCS61 — 2026-10-30
## Design
- GDMA1 vs Control, rerunning DESeq2 after removing the low-library control NCS61
  (1.6M reads; original control n=11)

## Results
- mtDNA anchor: 13/13 down, mean LFC=-1.37 (consistent with the main analysis -1.41 to -1.66)
  → anchor is robust
- DEG: 3,865 (up 1,333 / down 2,532) vs main analysis 3,950 (1,368/2,582)
- Overlap with the main-analysis significant set 3,563/3,950; Jaccard=0.838

## Conclusion (Act)
- Conclusions unchanged → the main analysis retains NCS61 (preserving control n=11 power); this
  sensitivity analysis is attached as supplementary material to the paper
