= Solution
Two useful changes would be:
* Show patient estimates together with matched, relevant cell-type controls and an independent measurement of modification at prespecified or held-out loci. Display uncertainty and all evaluated loci, or state a selection rule independent of the reported contrast. This separates genuine agreement from selection of high <xPore> values and from unrelated cell-background effects.
* Add performance across measured RNA-input levels for the same biological materials, with biological and technical replication, <read coverage> and failure rates reported. This tests the low-input part of the claim directly rather than inferring it from the heatmap.
A single additional knockout row would not address both independent accuracy and input dependence.
Back to article page
