I’m Justin Whalley at Rosalind Franklin University, corresponding author on a manuscript currently under review, “Multi-omic Glial Programs in Brain Aging” (Whalley J.P., Hunsberger H.C., Bennett D.A., Lamar M., Peterson D.A., Stutzmann G.E.). We generated derivative analysis outputs from existing ROSMAP modalities: snRNA-seq, DNA methylation, H3K9ac ChIP-seq, and TMT proteomics. Specifically, we applied Sparse Decomposition of Arrays (SDA) to the 276 DLPFC ROSMAP samples with snRNA-seq data and produced 546 latent components. We’d like to deposit the component-loading matrices and associated feature metadata alongside the upstream ROSMAP data, ideally wherever downstream analysis outputs are conventionally placed in the ROSMAP / AMP-AD Synapse structure. The proposed deposit is about 75 MB and currently includes: - 6 component-loading data files: - component × sample loadings - component × cell-type loadings - feature × component matrices for RNA, methylation, H3K9ac ChIP-seq, and proteins, with PIP ≤ 0.5 entries masked - 4 feature-metadata sidecars, one per feature matrix, decoding row labels with genomic coordinates and relevant annotations - 1 README / data dictionary describing file structure, feature IDs, sign convention, and metadata fields A companion GitHub repository with reproducible python notebooks for generating the manuscript figures from these matrices is currently private and will be made public on acceptance. Could you advise on the appropriate intake pathway for this type of derivative ROSMAP analysis output? We’re happy to follow DCA metadata validation or any project-standard submission process. Members of ROSMAP study leadership are co-authors and can confirm the data context if needed.