AD-FlowSAGE manuscript-supporting data, predictions, and stable directed gene regulatory network

This repository contains the manuscript-supporting data release for the study “AD-FlowSAGE: Adaptive Directional-Flow Graph Learning for Gene Regulatory Network Inference.” The deposited release includes the frozen model-input feature matrices, regulatory-reference records, SOURCE- and NODE-grouped split manifests, matched-input benchmark and ablation results, propagation-depth diagnostics, source-data tables underlying the reported figures and statistical comparisons, the 418,488-candidate directed prediction artifact, the final 15,000-edge stable directed gene regulatory network, annotated network nodes, the 37-edge high-confidence core, the 2,003 low-pH-supported relations, GraphML and SIF network exports, low-pH derived feature matrices, feature dictionaries, provenance records, and SHA256 manifests. The published low-pH RNA-seq component is associated with NCBI BioProject PRJNA1358764. An additional gene-level pH 4.0 24-h comparison was provided directly by the authors of the source study for use in the present research and was not part of the published transcriptomic analysis. The deposited package contains only the gene-level derived statistics and model-input features required to reproduce the AD-FlowSAGE low-pH feature block. Third-party raw materials are not redistributed. In particular, the deposit does not include original STRING or KEGG source files, the original author-provided low-pH workbook or raw count matrices, or third-party software source code. Such resources remain subject to their respective database, provider, and software terms. The author-generated data and documentation in this deposit are released under the Creative Commons Attribution 4.0 International license, subject to the third-party exclusions documented within the package. Release version: 1.0.0 Archive SHA256:1137c01df9b839466c91d1506b054fbe2b0898fc4ae1f5383064974f3f070764