Rooneyxu

生物信息学出版级数据可视化：火山图、热图、UMAP/t-SNE、violin、dotplot、Manhattan/GWAS、spatial、barplot 等；含简单 2–4 panel patchwork。务必在用户要画生信 plot/figure、DESeq2/scanpy/RNA-seq/scRNA-seq/spatial 结果出图、ComplexHeatmap/ggplot2/matplotlib 投稿导出（svg/png/pdf/600dpi）时使用。含 NMI pastel、色盲友好、adjustText/ggrepel、communication/publication 分层导出；默认 Python，提及投稿/出版/R/ComplexHeatmap 且未指定后端时先问 Python 还是 R。勿用于：非生信图表、只要表格不要图、plotly 交互 dashboard、hero panel 复杂投稿 figure（→ nature-figure）、pipeline/流程/示意图（→ GenerateImage）、英文润色/DA statement/PPT。

Extract final STRAND-compatible subcellular spatial transcriptomics metadata from a paper PDF, DOI/article URL, repository page, dataset landing page, approved downloaded data file, or optional topic search such as "pancreas subcellular-resolution spatial datasets". Use this skill when the user wants a Datasets.xlsx-style metadata table plus an auditable extraction report for MERFISH, seqFISH, Xenium, CosMx, STARmap, Molecular Cartography, or related subcellular spatial datasets. Also use it for upstream paper/dataset discovery before extraction, missing-field audits, download plans, QC/count-convention explanations, and paper-to-dataset curation runs. Do not use for generic paper summaries, review-only articles without reusable data, non-spatial scRNA-seq, bulk RNA-seq, or spot-level spatial data unless the user explicitly asks.