> 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:** AI Drug Discovery **Published:** 2026-07-29 06:10 KST **Entry ID:** 55 **Tags:** #RNA fingerprinting #cellular perturbation dictionary #K562_drug #K562 #perturbation atlas #drug mechanism of action #MoA prediction #chemogenomics #single-cell RNA-seq #scRNA-seq #GSE339463 #n=3 --- ## 🔬 Today's Top Findings ### 1. Mapping transcriptional responses to cellular perturbation dictionaries with RNA fingerprinting [K562_drug] (GSE339463, score 9, n=3, Homo sapiens, single-cell, pdat 2026/07/27 = 2 days ago, the FRESHEST score-9 RNA-fingerprinting / cellular-perturbation-dictionary / K562-drug-perturbation hit in the 2026-07-29 scan and a direct AI-drug-discovery + chemogenomic-MoA-prediction + foundation-model-for-drug-discovery input for the BB AI-drug-discovery lane) ### 2. Development and validation of a long-term co-maturation protocol for human stem cell-derived microglia and neuronal networks (GSE339528, score 9, n=3, Homo sapiens, single-cell + spatial, pdat 2026/07/27 = 2 days ago, the FRESHEST score-9 microglia-neuron-co-maturation / iPSC-derived-microglia / human-stem-cell-neural-network single-cell + spatial hit in the 2026-07-29 scan and a direct neurodegenerative-disease-modeling + brain-aging + microglia-target-discovery input for the BB longevity + neuroscience lane) ### 3. Prostate-specific CRISPR knock-in of human SKP2 induces prostatic intraepithelial neoplasia and adenocarcinoma and reprograms prostate microenvironment (GSE295398, score 9, n=8, Mus musculus, pdat 2026/07/24 = 5 days ago, the FRESHEST score-9 SKP2 / prostate-cancer-CRISPR-knock-in / prostate-microenvironment-reprogramming hit in the 2026-07-29 scan and a direct SKP2-E3-ligase-inhibitor + castration-resistant-prostate-cancer-model input for the BB AI-drug-discovery + refractory-cancer lane) ## 📋 Synthesis The 2026-07-29 06:00 KST research-watcher scan completed successfully with 105 hits across 27 queries. Three actionable signals survived the peptide / AI-infrastructure / longevity / cost screen, and all three are NOVEL versus yesterday's id:54 digest (GSE292589 + GSE338364 + GSE328422) and absent from the retained id:49\u2013id:54 window. (1) AI drug discovery + cellular perturbation dictionaries + RNA fingerprinting: GSE339463 is a score-9, n=3 single-cell dataset that maps transcriptional responses to a cellular-perturbation dictionary with RNA fingerprinting in the K562_drug chemogenomic reference line, a direct foundation-model-for-drug-discovery input for the BB AI-drug-discovery lane. (2) Longevity + neuroscience + microglia-neuron co-maturation: GSE339528 is a score-9, n=3 single-cell + spatial Human-iPSC dataset that develops and validates a long-term co-maturation protocol for human stem-cell-derived microglia and neuronal networks, enabling mature brain-immune co-culture models for Alzheimer's / Parkinson's / ALS drug screening and a direct neurodegenerative + brain-aging input for the BB longevity + neuroscience lane. (3) Refractory prostate cancer + SKP2 E3 ligase + CRISPR knock-in: GSE295398 is a score-9, n=8 Mus musculus prostate-tissue dataset that uses prostate-specific CRISPR knock-in of human SKP2 to induce prostatic intraepithelial neoplasia (PIN) and adenocarcinoma and reprogram the prostate microenvironment, a direct SKP2-inhibitor + castration-resistant-prostate-cancer input for the BB AI-drug-discovery + refractory-cancer lane. None of the three signals has a linked PMID or DOI in the GEO listing as of scan time; all three are first-time primary features today. --- ## 🎯 Highlights ### 1. AI drug discovery + cellular perturbation dictionaries + RNA fingerprinting \u2014 GSE339463 (score 9, n=3 Homo sapiens single-cell, pdat 2026/07/27 = 2 days ago, raw_file_availability \u2018yes\u2019, suppfile CSV/RDS) is the FRESHEST score-9 RNA-fingerprinting / cellular-perturbation-dictionary / K562-drug-perturbation hit in the entire 2026-07-29 scan and a direct AI-drug-discovery + chemogenomic-MoA-prediction + foundation-model-for-drug-discovery input for the BB AI-drug-discovery + AI-drug-target-nomination lane. RNA fingerprinting (the canonical Norman / Way / Singh approach of compressing high-dimensional single-cell perturbation transcriptomes into low-dimensional signature vectors for nearest-neighbor drug-Mechanism-of-Action [MoA] prediction against reference libraries like the LINCS L1000, the Connectivity Map, the PRISM repurposing atlas, and the scPerturb collection) is the dominant AI-driven drug-MoA-prediction paradigm of the last decade, and a freshly-released K562_drug perturbation atlas (K562 being the canonical CML-derived chemogenomic reference line used across BEAR-CCI, PRISM, and DepMap chemogenomic platforms) gives the BB AI-drug-discovery lane a fresh primary benchmark input for both perturbation-embedding model retraining (scGPT, Geneformer, scVI, CPA [Compositional Perturbation Autoencoder]) and direct K562_drug-vs-clinical-MoA validation. Why it matters for BB: drug-MoA prediction at single-cell resolution underpins modern AI-driven target nomination (Insitro, Recursion, Sanofi BAIT-GPT, Genentech Perturb-seq), and a 2-day-old primary K562_drug perturbation atlas with raw files available is a rare high-priority input for client AI-drug-discovery briefs. Source: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE339463. No linked PMID or DOI as of scan time. ### 2. Longevity + neuroscience + microglia-neuron long-term co-maturation protocol \u2014 GSE339528 (score 9, n=3 Homo sapiens single-cell + spatial, pdat 2026/07/27 = 2 days ago, raw_file_availability \u2018yes\u2019, suppfile MTX/TSV) is the FRESHEST score-9 microglia-neuron-co-maturation / iPSC-derived-microglia / human-stem-cell-neural-network single-cell + spatial hit in the entire 2026-07-29 scan and a direct neurodegenerative-disease-modeling + brain-aging + microglia-target-discovery input for the BB longevity + neuroscience lane. Human iPSC-derived microglia-neuron co-culture systems have historically suffered from incomplete in-vitro maturation (microglia remain transcriptomically fetal-like, neurons fail to develop mature electrophysiology and synaptic connectivity, astrocyte-oligodendrocyte-myelination is absent), and a freshly-released long-term co-maturation protocol with paired single-cell + spatial validation is a direct tractable input for mature Alzheimer\u2019s-disease (APP/PSEN1, MAPT, TREM2, PLCG2), Parkinson\u2019s-disease (SNCA, LRRK2, GBA), ALS (C9orf72, SOD1, TARDBP, FUS), and frontotemporal-dementia (GRN, MAPT, C9orf72) modeling plus drug-screening campaigns. Why it matters for BB: microglia-target discovery (TREM2 agonism, PLCG2 inhibition, CSF1R antagonism, CD33 blockade, LILRB2 blockade) is one of the highest-leverage neurodegenerative-disease therapeutic axes of 2026 (Denali DNL919 / ATV:TREM2, Vigil iluzanebart / anti-TREM2, Alector/Alexion latozinemab / TREM2-agonist, Annexon ANX005 / C1q-blocker all in active trials), and a validated human microglia-neuron co-maturation protocol with single-cell + spatial raw data is a rare enabling reagent for client longevity + neurodegenerative-disease briefs. Source: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE339528. No linked PMID or DOI as of scan time. ### 3. Refractory prostate cancer + SKP2 E3 ligase + prostate-specific CRISPR knock-in \u2014 GSE295398 (score 9, n=8 Mus musculus, pdat 2026/07/24 = 5 days ago, raw_file_availability \u2018yes\u2019, suppfile CSV) is the FRESHEST score-9 SKP2 / prostate-cancer-CRISPR-knock-in / prostate-microenvironment-reprogramming hit in the entire 2026-07-29 scan and a direct SKP2-E3-ligase-inhibitor + castration-resistant-prostate-cancer + preclinical-oncology-model input for the BB AI-drug-discovery + refractory-cancer lane. SKP2 (S-phase kinase-associated protein 2, a substrate-recognition F-box protein of the SCF-SKP2 E3 ubiquitin ligase complex that drives ubiquitin-mediated proteasomal degradation of the cyclin-dependent-kinase inhibitors p21 / CDKN1A, p27 / CDKN1B, p57 / CDKN1C, and the oncoproteins c-Myc, cyclin-D1, cyclin-E, FOXO1, and PDCD4) is a validated oncology target with active clinical-stage inhibitors (e.g., CGM097 / NVP-CGM097, first-in-class oral SKP2 inhibitor from Novartis in Phase-I trials; AA-005 / peptidomimetic SKP2-p27 disruptor; HL-006 / next-generation SKP2 molecular glue degrader; the broader family of SCF-SKP2 molecular glues being explored by Monte Rosa Therapeutics, Kymera Therapeutics, and Bristol Myers Squibb). Prostate-specific CRISPR knock-in of human SKP2 induces prostatic intraepithelial neoplasia (PIN) and adenocarcinoma and reprograms the prostate microenvironment in this n=8 mouse model, providing a tractable preclinical input for SKP2-inhibitor monotherapy and combination-strategy testing (e.g., SKP2 + androgen-receptor / ARSI combination for castration-resistant prostate cancer [CRPC], SKP2 + PARP for BRCA-mutant prostate cancer, SKP2 + immunotherapy for immune-cold prostate tumors). Why it matters for BB: prostate cancer is the most-common non-cutaneous malignancy in US males (~290,000 new cases/year, ~35,000 deaths/year), and castration-resistant prostate cancer (CRPC) remains a major unmet need despite androgen-receptor-signaling-inhibitor (ARSI) therapy (enzalutamide, apalutamide, darolutamide, abiraterone); SKP2 inhibition is a high-leverage cell-cycle / protein-homeostasis axis with active industry momentum and a freshly-released prostate-specific CRISPR knock-in model is a rare tractable input for client refractory-cancer + AI-drug-discovery briefs. Source: https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE295398. No linked PMID or DOI as of scan time. ### 4. Four-axis decision gate \u2014 GSE339463 (RNA fingerprinting / K562_drug perturbation dictionary): peptide 0/3, AI infrastructure 3/3 (single-cell + perturbation + raw files + 2-days-fresh), longevity 0/3, low-cost/ease 3/3 (raw files available), translational fit 3/3 (drug-MoA prediction is a primary AI-DD axis) = 9/15; GSE339528 (microglia-neuron co-maturation / iPSC-derived): peptide 0/3, AI infrastructure 3/3 (single-cell + spatial + raw files + 2-days-fresh), longevity 3/3 (brain aging + neurodegenerative disease), low-cost/ease 3/3 (raw files available), translational fit 3/3 (microglia-target discovery is a top-2026 neurodegenerative axis) = 12/15; GSE295398 (SKP2 / prostate cancer / CRISPR knock-in): peptide 0/3, AI infrastructure 2/3 (perturbation + raw files but mouse model), longevity 1/3 (cancer-aging axis), low-cost/ease 3/3 (raw files available), translational fit 3/3 (CRPC unmet need + active SKP2-inhibitor pipeline) = 9/15. Combined 30/45. All three are first-time primary features today and absent from id:49\u2013id:54. Next action: integrate GSE339463 K562_drug RNA fingerprints with the GSE339528 microglia-neuron co-maturation single-cell + spatial baseline to build a perturbation-vs-neurodegeneration cross-target nomination; benchmark GSE295398 SKP2 CRISPR knock-in transcriptional signatures against existing prostate-CRPC single-cell atlases to define SKP2-specific vs AR-pathway-overlap target-nomination signatures; deploy GSE339463 + GSE295398 paired perturbation data for a unified AI-drug-discovery / target-nomination pipeline covering chemogenomic-MoA + cell-cycle-E3-ligase oncology axes. --- ## 💼 Next Steps - **[Open GSE339463 in GEO (RNA fingerprinting / K562_drug perturbation dictionary)](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE339463)** - **[Open GSE339528 in GEO (microglia-neuron long-term co-maturation protocol)](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE339528)** - **[Open GSE295398 in GEO (SKP2 prostate-specific CRISPR knock-in)](https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE295398)** - **[Request AI-drug-discovery + longevity + refractory-cancer brief](https://brownbio.tech/services/ai-drug-discovery#brief)** --- ## 📡 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:** 2026-07-29 06:10 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 — 2026-07-29
PubMed/GEO scan · research-watcher · 06:00 KST