AlphaFold3 structural predictions for the protein-protein interactome of Prochlorococcus marinus subsp. pastoris str. CCMP1986 (MED4)

AlphaFold3 (AF3) structural predictions for the Prochlorococcus marinus MED4 protein-protein interactome. This dataset accompanies the manuscript "Topological entrenchment of adaptive proteins in the streamlined interactome of Prochlorococcus MED4" by Daakour, Nelson, Jaiswal et al. (in preparation). Contents 2,841 protein-pair AF3 predictions across three sets: CrY2H-seq positive pairs (Y2H, 1,084), random control pairs (RND, 1,083), and an additional curated set (ksa-ivenn, ~657). Each pair includes the top-ranked predicted complex (_model.cif), per-residue and pair-level confidence metrics (_confidences.json, _summary_confidences.json), and ranking scores (_ranking_scores.csv). Figure S7 source data (PPIs_AF3_FigS7 spreadsheet) and composite figure renders (FigureS7_composite.png and .pdf). Note on excluded files: Raw MSA input files (*_data.json, ~100 GB total) are excluded as they are regenerable AF3 inputs derived from public sequence databases, not scientific results. They are available from the corresponding authors on request. Linked records Raw CrY2H-seq sequencing reads: NCBI BioProject PRJNA1467408 (SRA Study SRP702555; Runs SRR38781112–SRR38781123). Link: https://www.ncbi.nlm.nih.gov/bioproject/PRJNA1467408 Top-ranked complexes: ModelArchive (accessions pending). Analysis code: GitHub (URL pending). Confidence metrics ipTM (interface predicted TM-score): >0.8 = high confidence; 0.6–0.8 = moderate; <0.6 = low. pLDDT: per-residue confidence (0–100). PAE (predicted aligned error, in Å): present in *_confidences.json. AlphaFold3 license: These structures were generated using AlphaFold3. Use is subject to the AlphaFold3 Model Parameters Terms of Use and Output Terms of Use — see TERMS_OF_USE.md files included with each prediction. Organism: Prochlorococcus marinus subsp. pastoris str. CCMP1986 (NCBI taxid 59919; RefSeq NC_005072.1). Funding: NYU Abu Dhabi Research Institute (grant AD060).