Controlled reanalysis of the nitroplast (UCYN-A) transit peptide: analysis code, derived tables and figures

Analysis code, derived tables and figures supporting the manuscript "The nitroplast import signal is a C-terminal, glycine- and alanine-rich, proline-punctuated element: a controlled reanalysis that separates real targeting grammar from cross-species composition artefact" (Boyraz C., submitted to Plant Molecular Biology). Published characterisations of the UCYN-A transit peptide (uTP) compare haptophyte sequences directly with plant chloroplast transit peptides, a design that confounds the targeting signal with proteome-wide amino-acid composition. This deposit contains the eight control layers built to remove that confound: a within-species contrast against internal windows of the same proteins (AUC 0.9425 against a positional null band of 0.5579-0.6274), homology filtering at 40 and 60 percent identity, species-blocked cross-validation, a profile-HMM circularity audit, comparison against plant mitochondrial transit peptides and secretory signal peptides, a positional null panel across nine organisms spanning four clades and three symbiotic states, an ortholog terminal length-shift test, and a window-length sensitivity sweep from 40 to 300 residues. All input data are public and are identified by accession; none is redistributed here. Full documentation, including the known reproducibility limits, is in README.md and code/MANIFEST.md. The analysis was executed in an agentic scientific-computing environment (Claude Science, Anthropic) under the author's direction; the author designed the controls, verified every reported value against its source table, and takes full responsibility for the content and conclusions.