Reproducibility data for the inverse-design DED-pipeline manuscript (PLOS Computational Biology revision)

Reproducibility data and code snapshot for the revised DED-pipeline manuscript (PLOS Computational Biology). inverse_design_paper_v1.2.tar.gz is split into eleven 100 MB parts (.part00 to .part10) because of upload size limits. Download all parts and join them with cat inverse_design_paper_v1.2.tar.gz.part* > inverse_design_paper_v1.2.tar.gz, then check the result against SHA256SUMS.txt. The archive contains particle sets with random-forest target-fit weights, summary metrics, target metrics, derived analysis tables, per-figure source data, and the ARCADE model jar. Its README.md describes the layout, SOURCES.tsv gives the origin of every file, and MANIFEST.SHA256 checksums the bundle. The random-forest weights are not calibrated posterior probabilities because the sampler omits the ABC-SMC proposal-density correction. inverse_design_code_plos_revision_6f4ba57.zip is a snapshot of the DED pipeline code and the revision analysis scripts (paper/plos_revision/, which maps each manuscript item to its script). SHA256SUMS.txt lists the checksums of the joined archive, each part, and the code snapshot. Supersedes v1.1.