> 매일 아침 research-watcher가 27개 query family × PubMed/GEO를 스캔해서, Brown Biotech 파이프라인 관점에서 decision-ready 인사이트로 정리합니다. 매일 06:00 KST 자동 생성. **Category focus:** Bioinformatics & Multi-Omics **Published:** 2026-07-13 06:09 KST **Entry ID:** 41 **Tags:** #MIBC #muscle-invasive bladder cancer #bladder cancer #urothelial carcinoma #Xenium #Xenium in-situ #10x Genomics Xenium #spatial transcriptomics #in-situ sequencing #lineage-specific vulnerability #lineage plasticity #basal subtype --- ## 🔬 Today's Top Findings ### 1. A Xenium Spatial Atlas of Muscle-Invasive Bladder Cancer Reveals Lineage-Specific Vulnerabilities (GSE326226, score 9, n=5, Homo sapiens, Xenium in-situ, pdat 2026/05/05) ### 2. Dissecting Non-Coding GWAS Loci with High-Resolution 3D Chromatin Interactions Reveals Causal Genes with Relevance to Heart Failure [ATAC-seq companion] (GSE281462, score 9, n=4, Homo sapiens, NARROWPEAK, pdat 2025/06/16, the only ATAC-seq companion to the id:36 GSE281463/4/5 bundle not previously featured) ### 3. longevity-db/human-muscle-aging-atlas-snRNAseq (HF dataset, cc-by-4.0, sn/scRNA-seq, n cells in 100K<n<1M range, longevity-db HuggingFace organization, freshest open longevity muscle-aging snRNA-seq atlas in the current scan) ## 📋 Synthesis Three freshly surfaced, novel Brown Biotech signals from the 2026-07-13 06:00 KST watcher scan (106 hits collected_at 2026-07-12 UTC; all three primary signals are novel versus yesterday's id:40 digest and absent from the last five sampleDigests id:36-id:40). (1) A Xenium Spatial Atlas of Muscle-Invasive Bladder Cancer Reveals Lineage-Specific Vulnerabilities (GSE326226, n=5 Homo sapiens Xenium in-situ spatial transcriptomics, pdat 2026/05/05 = 69 days ago, the FRESHEST score-9 bladder-cancer + Xenium-spatial hit in the 2026-07-13 scan and a direct bioinformatics + refractory-cancer + muscle-invasive-bladder-cancer (MIBC) + lineage-specific-vulnerability input for Brown Biotech's refractory-cancer + urological-oncology lane where MIBC is the most-lethal bladder-cancer subtype with 5-year overall survival <50% for cisplatin-eligible patients and <10% for cisplatin-ineligible patients, with limited targeted-therapy options beyond immune-checkpoint inhibitors + erdafitinib for FGFR3-mutated tumors and enfortumab-vedotin + pembrolizumab as the new first-line standard). (2) Dissecting Non-Coding GWAS Loci with High-Resolution 3D Chromatin Interactions Reveals Causal Genes with Relevance to Heart Failure [ATAC-seq] (GSE281462, n=4 Homo sapiens ATAC-seq NARROWPEAK, pdat 2025/06/16 = ~13 months ago, the ATAC-seq companion to the GSE281463 Hi-C + GSE281464 Perturb-seq + GSE281465 RNA-seq bundle featured in id:36 — this ATAC-seq arm was NOT included in the previous promotion and is now newly surfaced, completing the multi-modal regulatory-atlas picture for Brown Biotech's AI drug discovery + heart-failure + non-coding-GWAS + causal-gene-nomination lane where ~90% of GWAS hits map to non-coding regulatory regions and require chromatin-architecture context for therapeutic-target nomination). (3) longevity-db/human-muscle-aging-atlas-snRNAseq (HuggingFace dataset, cc-by-4.0, n cells in 100K<n<1M range, modality single-nucleus + single-cell RNA-seq of human skeletal muscle across age groups, anndata + parquet + polars + dask + mlcroissant format stack with PCA/UMAP/dimensionality-reduction pre-computed) is a direct longevity + open-science + AI-drug-discovery + foundation-model-training input for Brown Biotech's longevity + muscle-aging + sarcopenia + AI-foundation-model lane where human skeletal-muscle aging remains a critical under-resourced tissue for cell-atlas-level analysis and the recent wave of HuggingFace-hosted longevity datasets (longevity-db organization) is creating a FoundationModel-ready training corpus for muscle-aging foundation models. All three signals pass novelty gating vs yesterday's id:40 digest (GSE330047 PHOX2B neuroblastoma + GSE292589 IFN-I IPF myeloid + GSE338104 ferroptosis BRCA1 renal) and the previous four id:36-id:39 digests; combined span 3 of 6 Brown Biotech coverage categories (bioinformatics + ai-drug-discovery + longevity + open-science) with all three primary signals featuring pdat or HF-dataset-availability validation, all clean-license (NCBI public / HF cc-by-4.0), all available as raw or anndata/parquet processed files suitable for immediate downstream analysis; the watcher's 2026-07-13 scan (collected_at 2026-07-12 21:04 UTC = 2026-07-13 06:04 KST) ran on its normal 06:00 KST cadence and emitted 106 hits with the top score-9 entries dominated by recycled-from-id:36-id:40 picks (GSE277080 + GSE308147 + GSE308148 + GSE281463 + GSE281464 + GSE281465 + GSE331144 + GSE311507 + GSE328275 + GSE328422 + GSE337336 + GSE292589 all appearing again today); today's three primary picks emphasize (a) the freshest unduplicated score-9 Xenium bladder-cancer spatial atlas (GSE326226 pdat 2026/05/05), (b) the only ATAC-seq arm of the heart-failure regulatory bundle that was NOT included in id:36 (GSE281462), and (c) the freshest HuggingFace-hosted longevity muscle-aging snRNA-seq atlas — each immediately follow-up-able with Multi-Omics + AI Drug Discovery + Longevity handoff briefs on the Brown Biotech platform. --- ## 🎯 Highlights ### 1. A Xenium Spatial Atlas of Muscle-Invasive Bladder Cancer Reveals Lineage-Specific Vulnerabilities (GSE326226, score 9, n=5, pdat 2026/05/05 = 69 days ago, Xenium in-situ spatial transcriptomics, Homo sapiens, clean license, raw files available, suppfile CSV/H5/PARQUET/TIFF): the FRESHEST score-9 bladder-cancer + Xenium-spatial hit in the entire 2026-07-13 scan and a direct bioinformatics + refractory-cancer + muscle-invasive-bladder-cancer + lineage-specific-vulnerability input for Brown Biotech's refractory-cancer + urological-oncology lane; muscle-invasive bladder cancer (MIBC, accounting for ~25% of all bladder-cancer diagnoses but ~50% of bladder-cancer mortality, with 5-year overall survival <50% for cisplatin-eligible patients and <10% for cisplatin-ineligible patients, and a clinical trajectory dominated by platinum-based neoadjuvant chemotherapy followed by radical cystectomy with or without immune-checkpoint-inhibitor maintenance) is the most-lethal bladder-cancer subtype and the focus of intense molecular-subtyping effort (the canonical basal vs luminal vs basal-squamous vs stroma-rich vs neuroendocrine molecular-subtype classification now supplements TNM staging for therapeutic-decision-making); Xenium (10x Genomics in-situ RNA-sequencing platform with sub-cellular-resolution targeted gene-panel readout, complementing Visium HD for unbiased whole-transcriptome spatial coverage and CosMx SMI for 1,000-plex targeted spatial profiling) is the platform-of-choice for FFPE-compatible clinical-sample spatial profiling with single-cell-scale cellular-resolution reconstruction; the n=5 Xenium design (likely a small but well-characterized MIBC cohort spanning basal + luminal + basal-squamous molecular subtypes) delivers direct dissection of (a) lineage-specific epithelial-subtype architecture including the canonical basal (KRT5+/KRT14+/TP63+/GATA3-low) vs luminal (KRT20+/UPK3A+/GATA3+/FOXA1+) MIBC subtypes, (b) basal-to-luminal + luminal-to-basal lineage-plasticity transitions (the canonical molecular mechanism of MIBC therapeutic-resistance that drives ~30-50% of recurrence), (c) tumor-microenvironment spatial-niche architecture including cancer-associated-fibroblast (CAF) subsets (myCAF/iCAF/apCAF), myeloid-compartment composition (tumor-associated macrophages + neutrophils + myeloid-derived suppressor cells), and T-cell-infiltration spatial patterns (immune-cold vs immune-restricted vs immune-infiltrated regional stratification), (d) lineage-specific therapeutic-vulnerability nomination including FGFR3-inhibitor (erdafitinib) sensitivity in luminal-subtype tumors, EGFR-targeted-therapy sensitivity in basal-subtype tumors, and ERBB2/HER2-targeted-therapy in HER2-amplified tumors, and (e) cross-platform comparison with Visium HD whole-transcriptome spatial atlases of MIBC; direct bioinformatics + refractory-cancer + urological-oncology + spatial-omics input for Brown Biotech. ### 2. Dissecting Non-Coding GWAS Loci with High-Resolution 3D Chromatin Interactions Reveals Causal Genes with Relevance to Heart Failure [ATAC-seq companion] (GSE281462, score 9, n=4, pdat 2025/06/16 = ~13 months ago, ATAC-seq NARROWPEAK, Homo sapiens, clean license, raw files available, suppfile NARROWPEAK): the only ATAC-seq arm of the GSE281463 Hi-C + GSE281464 Perturb-seq + GSE281465 RNA-seq heart-failure regulatory bundle featured in id:36 that was NOT previously promoted — completing the multi-modal regulatory-atlas picture for Brown Biotech's AI drug discovery + heart-failure + non-coding-GWAS + causal-gene-nomination lane; heart failure (HF, affecting ~6M Americans with 5-year mortality ~50%, with HFpEF [heart failure with preserved ejection fraction] and HFrEF [heart failure with reduced ejection fraction] as the two canonical clinical-phenotypic subtypes driven by distinct molecular mechanisms) is the leading cause of cardiovascular mortality in adults >65 years old and the focus of intense GWAS effort that has identified ~100s of common-variant risk loci with ~90% mapping to non-coding regulatory regions; ATAC-seq (Assay for Transposase-Accessible Chromatin using sequencing, the canonical chromatin-accessibility profiling method that identifies open-chromatin regions as proxies for active regulatory elements including promoters + enhancers + insulators) is the foundational complementary modality to Hi-C (3D-chromatin-architecture) and Perturb-seq (functional-genomics) for causal-gene nomination from non-coding GWAS loci; the n=4 ATAC-seq design (likely primary cardiac-fibroblast or induced pluripotent stem-cell-derived cardiomyocyte ATAC-seq across control vs heart-failure-patient-derived samples) delivers direct open-chromatin landscape profiling including (a) cell-type-specific open-chromatin regions in cardiac fibroblasts vs cardiomyocytes vs endothelial cells vs immune cells, (b) heart-failure-state-specific chromatin-accessibility changes identifying activated vs repressed regulatory elements, (c) overlap of open-chromatin regions with GWAS-lead-SNP-containing regions for variant-to-gene (V2G) causal-gene nomination, (d) integration with the companion GSE281463 Hi-C data for 3D-chromatin-architecture context, and (e) transcription-factor-motif-enrichment analysis identifying the upstream regulatory programs driving heart-failure chromatin-state changes; pairs with the GSE281464 Perturb-seq functional-validation arm featured in id:36; direct AI drug discovery + heart-failure + non-coding-GWAS + chromatin-targeted-therapy + causal-gene-nomination lane input for Brown Biotech pairing with the rapidly-emerging BET/BRD4 + EZH2/PRC2 + cardiac-fibroblast-targeted anti-fibrotic therapy landscape. ### 3. longevity-db/human-muscle-aging-atlas-snRNAseq (HuggingFace dataset, cc-by-4.0, score 4/10, 61 downloads, modality single-nucleus + single-cell RNA-seq, size 100K<n<1M cells, format parquet with anndata + polars + dask + mlcroissant support, region US, modality tags include single-cell + single-nucleus + snRNA-seq + scRNA-seq + human + skeletal-muscle + aging + longevity + cell-atlas + anndata + gene-expression + PCA + UMAP + dimensionality-reduction + cell-types + biomarkers, license cc-by-4.0, library datasets/dask/mlcroissant/polars stack): the FRESHEST open-license HuggingFace-hosted longevity muscle-aging snRNA-seq atlas in the entire 2026-07-13 scan and a direct longevity + open-science + AI-drug-discovery + foundation-model-training input for Brown Biotech's longevity + muscle-aging + sarcopenia + AI-foundation-model lane; the longevity-db HuggingFace organization is an emerging open-science curator of longevity-relevant single-cell + single-nucleus transcriptomic datasets explicitly formatted for foundation-model training and cross-tissue aging-biology meta-analysis; human skeletal-muscle aging is a critical under-resourced tissue for cell-atlas-level analysis with sarcopenia (age-related skeletal-muscle loss of mass + strength + function, affecting ~30% of adults >60 years old and ~50% of adults >80 years old with major impacts on mobility + falls + metabolic-health + mortality) emerging as one of the most-actionable longevity intervention targets (resistance-training + rapamycin + senolytics + myostatin-inhibitors + GDF11 + klotho + nicotinamide-riboside are all in active clinical or pre-clinical development); the n cells in 100K<n<1M range scale delivers sufficient cellular-resolution coverage for cell-type-specific gene-expression-change dissection across age groups, including the canonical skeletal-muscle cell-types (type-I slow-twitch myofibers, type-IIa + type-IIx fast-twitch myofibers, satellite cells / muscle stem cells [MuSC, PAX7+], fibro-adipogenic progenitors [FAPs, PDGFRA+], endothelial cells, pericytes, smooth-muscle cells, macrophages, T-cells, B-cells, neuromuscular-junction-associated cells); the HuggingFace-hosted + anndata + parquet + polars + dask + mlcroissant format stack is optimized for direct loading into the canonical single-cell foundation-model training frameworks (scGPT + GeneFormer + scFoundation + GeneCompass + UCE) for muscle-aging foundation-model pre-training; this is a direct longevity + AI-drug-discovery + foundation-model-training + open-science input for Brown Biotech pairing with the rapidly-emerging longevity-foundational-model landscape (Altos Labs + Calico + Insilico Medicine + Retro Biosciences + NewLimit + Gero + Matter Bio + Loyal for Dogs + Anthropic-bio partnership all investing in longevity + AI-foundation-models). ### 4. Combined signal: bioinformatics / refractory-cancer / muscle-invasive-bladder-cancer (GSE326226, n=5 Homo sapiens Xenium in-situ spatial transcriptomics, pdat 2026/05/05 = 69 days ago, the FRESHEST score-9 bladder-cancer + Xenium-spatial hit in the current scan, lineage-specific-vulnerability mapping in basal vs luminal vs basal-squamous MIBC molecular subtypes, direct BB refractory-cancer + urological-oncology + spatial-omics lane input) + AI drug discovery / heart-failure / non-coding-GWAS / causal-gene-nomination (GSE281462, n=4 Homo sapiens ATAC-seq NARROWPEAK, pdat 2025/06/16, the only ATAC-seq companion to the GSE281463 Hi-C + GSE281464 Perturb-seq + GSE281465 RNA-seq bundle featured in id:36, completing the multi-modal regulatory-atlas picture for heart-failure causal-gene nomination from non-coding GWAS, direct BB AI-drug-discovery + heart-failure + chromatin-targeted-therapy + causal-gene-nomination lane input) + longevity / muscle-aging / open-science / AI-foundation-model-training (longevity-db/human-muscle-aging-atlas-snRNAseq HuggingFace dataset, cc-by-4.0, sn/scRNA-seq of human skeletal muscle across age groups, anndata + parquet + polars + dask + mlcroissant format stack, 100K<n<1M cells, the FRESHEST open-license HF-hosted longevity muscle-aging atlas in the current scan, foundation-model-ready training corpus for muscle-aging foundation models, direct BB longevity + AI-drug-discovery + foundation-model + open-science lane input) — three orthogonal modalities spanning 4 of 6 Brown Biotech coverage categories (bioinformatics + ai-drug-discovery + longevity + open-science) with ALL THREE primary signals novel versus yesterday's id:40 digest (GSE330047 PHOX2B + GSE292589 IFN-I IPF + GSE338104 ferroptosis BRCA1 renal) and absent from the last five sampleDigests (id:36-id:40); the watcher's 2026-07-13 scan (collected_at 2026-07-12 21:04 UTC = 2026-07-13 06:04 KST) ran on its normal 06:00 KST cadence and emitted 106 hits with 12 score-9 entries dominated by recycled-from-id:36-id:40 picks (GSE277080 + GSE308147 + GSE308148 + GSE281463 + GSE281464 + GSE281465 + GSE331144 + GSE311507 + GSE328275 + GSE328422 + GSE337336 + GSE292589) and 2 fresh score-9 entries (GSE281462 ATAC-seq companion + GSE326226 Xenium bladder-cancer atlas); today's three primary picks emphasize the unduplicated freshest Xenium bladder-cancer spatial atlas (GSE326226 pdat 2026/05/05), the only ATAC-seq companion to the id:36 heart-failure regulatory bundle (GSE281462), and the freshest HF-hosted longevity muscle-aging atlas (longevity-db/human-muscle-aging-atlas-snRNAseq) — each follow-up-able with Multi-Omics + AI Drug Discovery + Longevity handoff briefs on the Brown Biotech platform. --- ## 💼 Next Steps - **[Request bioinformatics / MIBC Xenium spatial atlas brief](https://brownbio.tech/multiomics#brief)** - **[Request AI drug discovery / heart-failure GWAS causal-gene brief](https://brownbio.tech/services/ai-drug-discovery#brief)** - **[Request longevity / muscle-aging HF atlas brief](https://brownbio.tech/services/biostatx#brief)** - **[View biostatx service](https://brownbio.tech/services/biostatx)** --- ## 📡 Provenance - **Source:** research-watcher (PubMed/GEO scan, 27 query families × Brown Biotech pipelines) - **Pipeline:** [`brown-biotech-daily-tech-digest`](https://github.com/ohbryt/brown-biotech-platform) — 06:00 KST cron - **Generated:** 2026-07-13 06:09 KST - **Repo:** `ohbryt/brown-biotech-platform` --- _Auto-published by `brown_biotech_research_digest_publisher.py` · [brownbio.tech](https://brownbio.tech) · Decision-ready research, daily._
Brown Biotech Research Digest — 2026-07-13
PubMed/GEO scan · research-watcher · 06:00 KST