Research Pulse — 2026-10-17 | Brown Biotech

# Research Pulse — 2026-10-17
_Generated 2026-10-17 07:30 KST · AI × Bioresearch Daily synthesis_

Brown Biotech daily research pulse — multi-front synthesis across CRISPR /
gene editing, AI drug discovery, protein AI, single-cell foundation models,
FDA/regulatory, and spatial transcriptomics. Sources: company filings,
peer-reviewed journals, FDA guidance, conference disclosures.

---

### Post 1: 오늘의 전체 흐름

🧬 **AI × Bioresearch Daily — 10월 17일 (금요일) · Clinicalizing(임상실현) Day**

10/17(임상실현) → 10/16(연결) → 10/15(규모화) → 10/14(평가) → 10/13(표준화) → 10/12(종합) → 10/11(문맥화) → 10/10(실현). Interfacing(10/16)된 Framework들이 **각 분야主流 체계에서 실제 임상 결과(Clinical Results)로 검증되는 단계로 진입하는 날** — Shaik et al.로 SCD gene therapy systematic review: 148 patients 9 programs + exa-cel VF12 96.7% + lovo-cel severe VOE 100% + delivery 44/46 exa-cel + 35/47 lovo-cel(European Journal Haematology 2026), Cohn et al.로 Lonvoguran ziclumeran Phase 3 RCT: HAE 单一回静脉输液 + attack rate -87% (P<0.001) + adverse events 92% + no serious/grade3 + NEJM 2026, Sheikh et al.로 Kresladi FDA approved March 2026: marnetegra-gene autotemcel + LAD-I + ITGB2 + March 26 2026(Annals Medicine Surgery 2026), Mao et al.로 AI drug discovery clinical translation: Phase IIa encouraging efficacy + Phase I safety signals + multimodal AI + digital twins + explainable AI(Frontiers Pharmacology 2026), Garcia et al.로 AlphaFold zero-shot de novo design: moderate discrimination + ESMFold pLDDT best + confidence topology-dependent + 614 benchmark monomers(Protein Science 2026), Li et al.로 perturbation benchmark: 13 methods 25 datasets + PerturbNet superior DE + no method cross cell-type + conservative bias + foundation models compress variance(Science Advances 2026), Shahzad et al.로 FDA CGT 38 approvals: 86.8% orphan + 92.1% expedited + 73.7% single study + 44.7% surrogate + 17.9% primary efficacy postmarketing(Clinical Pharmacology Therapeutics 2026), Dimitrov et al.로 single-cell interpretation review: causal inference + mechanistic discovery + disentanglement + population tracing + unifying ontology(Nature Reviews Genetics 2026)이 동시에 진행된다.

Clinicalizing은 Interfacing의 다음 단계다. 오늘 6개 전선은 연결된 Framework들이 **실제 임상 결과·실제 regulatory approval ·실제 experimental validation로 이행(Clinicalizing)되는 날**이다.

*출처: [Shaik SCD Gene Therapy European Journal Haematology 2026](https://consensus.app/papers/details/f71effec863b55a39d2a9333578ff946/), [Cohn Lonvoguran Ziclumeran NEJM 2026](https://consensus.app/papers/details/014f88be1af55665a357eb28a20d5cbc/), [Sheikh Kresladi FDA LAD-I Annals Medicine Surgery 2026](https://consensus.app/papers/details/f2f27f7392e05d9693a11da0cb87b29d/), [Mao AI Drug Discovery Frontiers Pharmacology 2026](https://consensus.app/papers/details/ca5321af0bb656e095b6108766231764/), [Garcia AlphaFold Zero-Shot Protein Science 2026](https://consensus.app/papers/details/f9c8c442c8065c0688de2538546747db/), [Li Perturbation Benchmark Science Advances 2026](https://consensus.app/papers/details/c9942dcafc665fd59f98c510a131e7e8/), [Shahzad FDA CGT 38 Approvals Clinical Pharmacology Therapeutics 2026](https://consensus.app/papers/details/24705d5037e85bcb9d9863f1aabca040/), [Dimitrov Single-Cell Interpretation Nature Reviews Genetics 2026](https://consensus.app/papers/details/4c9a5a9396a85b9797d6e1f0ecd4286b/)*

*#ClinicalizingDay #Shaik #EuropeanJournal #SCD #GeneTherapy #148Patients #9Programs #ExaCel #LovoCel #VF12 #VOE #Cohn #NEJM #Lonvoguran #Ziclumeran #HAE #Phase3 #87Percent #Sheikh #Kresladi #FDA #LADI #ITGB2 #Marnetegrave #March2026 #Mao #FrontiersPharmacology #AIDrugDiscovery #PhaseIIa #PhaseI #MultimodalAI #DigitalTwins #Garcia #ProteinScience #AlphaFold #ZeroShot #DeNovo #ESMFold #pLDDT #Li #ScienceAdvances #Perturbation #PerturbNet #13Methods #25Datasets #Shahzad #ClinicalPharmacology #38CGT #Orphan #Expedited #SingleStudy #Surrogate #Dimitrov #NatureReviews #Genetics #CausalInference #Mechanistic #UnifyingOntology #ClinicalResults #ClinicalTranslation #BiomedicalAI #DailyReport*

---

### Post 2: CRISPR & Gene Editing — Lonvoguran Ziclumeran Phase 3 × Kresladi 2026 Approval: Clinicalizing

🪄 **[CRISPR/Gene Editing] Shaik et al.로 SCD gene therapy systematic review: 148 patients 9 programs + exa-cel VF12 96.7% + lovo-cel severe VOE 100% + delivery 44/46 exa-cel + 35/47 lovo-cel(European Journal Haematology 2026) + Cohn et al.로 Lonvoguran ziclumeran Phase 3 RCT: HAE single-dose IV + attack rate -87% (P<0.001) + 92% adverse events + no serious/grade3 + NEJM 2026 + Sheikh et al.로 Kresladi FDA approved March 2026: marnetegra-gene autotemcel + LAD-I + ITGB2 + March 26 2026(Annals Medicine Surgery 2026): CRISPR Clinicalizing(임상실현) 동시 확인**

CRISPR gene therapy의 **Lonvoguran ziclumeran Phase 3 RCT + Kresladi 2026 approval + SCD systematic review clinical outcomes(Clinicalizing)**가 출처에서 동시에 확인된다[1][2][3].

**Lonvoguran Ziclumeran Phase 3 RCT: HAE -87% Attack Rate + NEJM[2]**
David Cohn et al.(New England Journal of Medicine 2026)의 연구는 Lonvoguran ziclumeran(Lonvo-z)의 Phase 3 임상을 보고한다[2]. **Phase 3 double-blind trial**: hereditary angioedema(HAE) 환자 80명, 2:1 비율. **Single intravenous infusion**: 50mg 단일 투여. **Primary endpoint**: investigator-confirmed HAE attacks/month (week 5–28). **Least-squares mean monthly attack rate**: Lonvo-z 0.26 vs placebo 2.10 — **relative difference -87%, P<0.001**. **Adverse events**: Lonvo-z 92%, placebo 86% (day 1–week 28). **No serious or grade 3 adverse events in Lonvo-z group**: 안전성 확인. **Most common >10%**: infusion-related reaction, headache, fatigue, back pain, upper respiratory tract infection. **HAELO ClinicalTrials.gov NCT06634420**: prospective validation. 핵심: Lonvoguran ziclumeran — CRISPR Phase 3 RCT 임상실현(Clinicalizing).

**SCD Gene Therapy Systematic Review: 148 Patients + Delivery Outcomes[1]**
M. Shaik et al.(European Journal of Haematology 2026)의 systematic review는 SCD gene therapy의 임상 결과를 종합한다[1]. **25 reports 9 programs 148 infused patients**: lentiviral, shmiR-BCL11A, CRISPR-Cas9, CRISPR-Cas12a, base editing platforms. **exa-cel VF12 96.7%** (29/30 pivotal cohort): vaso-occlusive event-free. **lovo-cel severe VOE resolution 100%** (25/25 evaluable): severe VOE 해결. **Both products achieved transfusion independence**: 두 제품 모두 수혈 독립 달성. **Delivery outcomes**: exa-cel 44/46 infused vs lovo-cel 35/47 infused — 제조/투여 격차 확인. **Long-term safety, fertility, organ outcomes, reimbursement, global scalability remain unresolved**: 이행의 과제. 핵심: SCD gene therapy — 이행의 delivery outcomes 격차 포함 임상실현(Clinicalizing).

**Kresladi FDA Approved March 2026: LAD-I First Gene Therapy[3]**
M. Sheikh et al.(Annals of Medicine and Surgery 2026)의 연구는 Kresladi(marnetegra-gene autotemcel)의 FDA 승인 사실을 보고한다[3]. **Kresladi FDA approved March 26, 2026**: leukocyte adhesion deficiency type I(LAD-I) 치료제. **ITGB2 gene mutation**: autosomal recessive combined immunodeficiency. **75% mortality by age 2**: 치료되지 않으면 높은 사망률. **Matched sibling donor 없는 환자를 위한 치료**: 수혜자 확대. **marnetegra-gene autotemcel mechanism**: 유전자 치료제. **Allogeneic HSCT 대안**: graft-versus-host disease 위험 회피. 핵심: Kresladi FDA approval — 2026년 CGT 임상실현(Clinicalizing).

**10월 16일 Interfacing(Lushington first-in-human + Feierman FDA framework + Ortega Moreno global challenges) vs 오늘의 차별점**: 10/16이 **Lushington Molecular Therapy**으로 prime editing first-in-human functional restoration을, **Feierman AJHG**으로 FDA personalized prime editing framework를, **Ortega Moreno International Science Journal**으로 global clinical translation challenges를 제시했다면, 오늘은 **Cohn NEJM**으로 **Lonvoguran ziclumeran Phase 3 RCT — -87% attack rate (P<0.001) + no serious adverse events + NEJM publication**이라는 CRISPR 최초 Phase 3 RCT 임상 결과(Clinicalizing)를, **Shaik European Journal**으로 **148 patients 9 programs + VF12 96.7% + VOE 100% + delivery 44/46 vs 35/47**이라는 delivery outcomes 격차 포함 현실(Clinicalizing)을, **Sheikh Annals**으로 **Kresladi March 2026 FDA approval + LAD-I + ITGB2**이라는 2026년 regulatory approval(Clinicalizing)을 동시에 제시한다. Interfacing(10/16) → Phase 3 RCT results + delivery outcomes + 2026 regulatory approval Clinicalizing(10/17).

실무 함의: 유전자 치료 개발팀에서 Cohn et al.의 Lonvoguran Phase 3 결과를 in vivo CRISPR clinical development strategy에 반영해야 한다.

*출처: [Shaik SCD Gene Therapy Systematic Review European Journal Haematology 2026](https://consensus.app/papers/details/f71effec863b55a39d2a9333578ff946/), [Cohn Lonvoguran Ziclumeran Phase 3 NEJM 2026](https://consensus.app/papers/details/014f88be1af55665a357eb28a20d5cbc/), [Sheikh Kresladi LAD-I FDA Approval Annals Medicine Surgery 2026](https://consensus.app/papers/details/f2f27f7392e05d9693a11da0cb87b29d/)*

*#CRISPR #GeneEditing #Shaik #EuropeanJournal #Haematology #SCD #GeneTherapy #148Patients #9Programs #ExaCel #LovoCel #VF12 #VOE #TransfusionIndependence #DeliveryOutcomes #Cohn #NEJM #Lonvoguran #Ziclumeran #HAE #Phase3 #87Percent #PValue #Infusion #Sheikh #Kresladi #FDA #Approved #March2026 #LADI #ITGB2 #Marnetegrave #Autotemcel #ClinicalizingDay #Phase3Results #RegulatoryApproval #ClinicalTranslation #BiomedicalAI #DailyReport*

---

### Post 3: AI 신약 — AI Clinical Candidates Phase IIa/Phase I 결과 + Multimodal AI: Clinicalizing

💊 **[AI 신약] Mao et al.로 AI drug discovery clinical translation: Phase IIa encouraging efficacy + Phase I safety signals + multimodal AI + digital twins + explainable AI(Frontiers Pharmacology 2026) + Seyhan et al.로 AI drug discovery review: scaffold-aware tools + 0건 완전 마케팅 승인 + AI clinical candidates 진입(Critical Reviews Oncology Hematology 2026) + Cai et al.로 preclinical AI pipeline review: multimodal AI + digital twins + closed-loop automation + 2020-2026 literature(Pharmaceuticals 2026): AI Drug Discovery Clinicalizing(임상실현) 동시 확인**

AI-Driven Drug Discovery의 **Phase IIa efficacy + Phase I safety + multimodal/digital twins clinical translation(Clinicalizing)**이 출처에서 동시에 확인된다[4][5][6].

**AI Clinical Translation: Phase IIa Encouraging + Phase I Safety Signals[4]**
Zihan Mao et al.(Frontiers in Pharmacology 2026)의 review는 AI drug discovery의 임상 전환 현황을 분석한다[4]. **One AI-discovered candidate Phase IIa with encouraging efficacy**: 실질적 효능 확인. **Another stalled in Phase I when safety signals surfaced**: 안전성 문제로 중단. **Double-edged story**: AI drug discovery 임상 이행 현실. **Multimodal AI integration**: 다양한 데이터 통합. **Digital twins**: 약물 개발에서의 디지털 트윈 적용. **Explainable AI for transparency and regulatory confidence**: 규제信心 확보를 위한 해석 가능 AI. **Challenges**: biological complexity + patchy data + shortage of ML-pharmaceutical scientists. **Traditional Chinese Medicine and natural product screening**: 새로운 적용 영역. 핵심: AI drug discovery — 임상 결과의 이중성(Clinicalizing).

**AI Drug Discovery Review 재확인: Scaffold-Aware + 0건 완전 승인 + Clinical Candidates 진입[5]**
(A. Seyhan et al. Critical Reviews in Oncology/Hematology 2026). scaffold-aware tools(AlphaFold, MoleR, PocketCrafter) + AI-originated candidates clinical development 진입 + 0건 완전 마케팅 승인. 핵심: AI drug discovery — 이행의 clinical candidates(Clinicalizing).

**Preclinical AI Pipeline Review: Multimodal + Digital Twins + Closed-Loop Automation[6]**
Meng Cai et al.(Pharmaceuticals 2026)의 review는 preclinical AI pipeline을 종합한다[6]. **Drug-target interaction prediction**: 표적 상호작용 예측. **Structure prediction**: 구조 예측. **De novo design**: de novo 분자 설계. **Virtual screening**: 가상 스크리닝. **Drug repurposing**: 약물 재창출. **ADMET forecasting**: 흡수/분포/대사/배설/독성 예측. **Multimodal AI**: 다양한 모달리티 통합. **Digital twins**: 약물 개발 디지털 트윈. **Closed-loop automation**: 폐쇄형 자동화. **2020-2026 literature**: 최신 동향. 핵심: Preclinical AI pipeline — 이행의 기술 체계(Clinicalizing).

**10월 16일 Interfacing(Huynh AI agents + He ChatInvent + Seyhan clinical candidates) vs 오늘의 차별점**: 10/16이 **Huynh Drug Discovery Today**으로 AI agents drug discovery 개념을, **He Drug Discovery Today**으로 AstraZeneca ChatInvent deployment를, **Seyhan 재확인**으로 0건 완전 마케팅 승인을 제시했다면, 오늘은 **Mao Frontiers Pharmacology**으로 **Phase IIa encouraging efficacy + Phase I safety signals라는 구체적 임상 결과 데이터**(Clinicalizing)를, **Cai Pharmaceuticals**으로 **multimodal AI + digital twins + closed-loop automation의 preclinical pipeline 체계**(Clinicalizing)를, **Seyhan 재확인**으로 ** scaffold-aware tools + clinical candidates 진입 현황**(clinical candidates Clinicalizing)을 동시에 제시한다. Interfacing(10/16) → Phase IIa/Phase I 구체적 임상 결과 + preclinical pipeline 체계 Clinicalizing(10/17).

실무 함의: AI drug discovery 협업팀에서 Mao et al.의 Phase IIa/Phase I 결과를 AI clinical development strategy에 반영해야 한다.

*출처: [Mao AI Drug Discovery Clinical Translation Frontiers Pharmacology 2026](https://consensus.app/papers/details/ca5321af0bb656e095b6108766231764/), [Seyhan AI Drug Discovery Critical Reviews Oncology Hematology 2026](https://consensus.app/papers/details/0e6b685fd2be56fb8a814115873a73d5/), [Cai Preclinical AI Pipeline Pharmaceuticals 2026](https://consensus.app/papers/details/9db7efc33cb250c4b08b448e73fcdd17/)*

*#AI #DrugDiscovery #Mao #FrontiersPharmacology #ClinicalTranslation #PhaseIIa #PhaseI #SafetySignals #MultimodalAI #DigitalTwins #ExplainableAI #Seyhan #CriticalReviews #OncologyHematology #ScaffoldAware #AlphaFold #MoleR #PocketCrafter #ZeroApproval #AIClinicalCandidates #Cai #Pharmaceuticals #Preclinical #DrugTarget #StructurePrediction #DeNovo #VirtualScreening #ADMET #ClosedLoop #Automation #ClinicalizingDay #ClinicalResults #TranslationalMedicine #BiomedicalAI #DailyReport*

---

### Post 4: 단백질 AI — AlphaFold Zero-Shot De Novo Design Validation + ProteinMPNN: Clinicalizing

🔬 **[단백질 AI] Garcia et al.로 AlphaFold zero-shot de novo design: 614 benchmark monomers + moderate discrimination + ESMFold pLDDT best + topology-dependent confidence + logistic regression modest improvement(Protein Science 2026) + Feldman et al.로 AlphaFold3 adversarial 40% invariance 재확인: confidence failures + template memorization(Computational Structural Biotechnology 2026) + Mi et al.로 Protein Design AI era review: AlphaFold2 + transformer + de novo design + therapeutic protein + enzyme catalysis(Computational Structural Biotechnology 2026): 단백질 AI Clinicalizing(임상실현) 동시 확인**

단백질 AI의 **AlphaFold zero-shot de novo design validation + ProteinMPNN discrimination 한계(Clinicalizing)**가 출처에서 동시에 확인된다[7][8][9].

**AlphaFold Zero-Shot De Novo Design: Moderate Discrimination + ESMFold pLDDT Best[7]**
Mario Garcia et al.(Protein Science 2026)의 연구는 AlphaFold의 de novo protein design에서의 zero-shot prediction 능력을 평가한다[7]. **614 experimentally characterized de novo monomers benchmark**: 11개 design studies 2012–2021. **All predictive models moderate ability to discriminate**: 실험적 성공/실패 판별 능력 중간. **Many failed designs have better confidence metrics than successful designs**: 역설적 결과. **Confidence metrics topology-dependent**: 토폴로지 의존적. **ESMFold average pLDDT best individual performance**: ESMFold pLDDT 최고 개별 성능. **Logistic regression combining all metrics modest improvement**: 앙상블 효과 제한적. **Filtering strategy prior to experimental validation**: 실험 전 필터링 전략으로는 유용하나 정확도 제한. 핵심: AlphaFold zero-shot de novo design — 이행의 판별 한계(Clinicalizing).

**AlphaFold3 Adversarial Evaluation 40% Invariance 재확인[8]**
(Jonathan Feldman et al. Computational Structural Biotechnology Journal 2026 — 10/16에서 상세 기술). 40% residue invariance + confidence metric unreliability 35% + template memorization vs biophysical reasoning. 핵심: AlphaFold3 adversarial limits — 이행의 신뢰성 한계(Clinicalizing).

**Protein Design AI Era Review: Transformer + De Novo + Therapeutic[9]**
Yan-Lin Mi et al.(Computational Structural Biotechnology Journal 2026)의 review는 AI-driven protein design의 전체 그림을 제시한다[9]. **AlphaFold2 near-experimental accuracy**: 구조 예측 Revolutionary. **Transformer-based protein language models**: de novo sequence design. **Therapeutic protein engineering**: 치료 단백질 공학. **Enzyme catalysis**: 효소 촉매. **Synthetic biology**: 합성생물학. **In silico to experimental validation pipeline**: 전임상 파이프라인. **Challenges**: model interpretability + training data biases + experimental validation rates. 핵심: Protein design AI era — 이행의 전임상 체계(Clinicalizing).

**10월 16일 Interfacing(Feldman AF3 adversarial + Mahtha ProModEv + Yin AF3 evolution) vs 오늘의 차별점**: 10/16이 **Feldman Computational Structural Biotechnology**으로 AF3 adversarial 40% invariance를, **Mahtha NAR Genomics Bioinformatics**으로 AF2/AF3 vs ESMFold accuracy benchmarking을, **Yin Frontiers Molecular Biosciences**으로 AF3 architecture evolution을 제시했다면, 오늘은 **Garcia Protein Science**으로 **614 benchmark monomers + moderate discrimination + ESMFold pLDDT best + topology-dependent confidence + failed designs have better metrics**라는 de novo design validation의 구체적 한계(Clinicalizing)를, **Feldman 재확인**으로 **40% invariance + confidence failures + template memorization의 한계 재확인**(clinical validation 한계 Clinicalizing)을, **Mi Computational Structural Biotechnology**으로 **AlphaFold2 + transformer + de novo + therapeutic protein + enzyme + synthetic biology의 전체 체계**(therapeutic protein Clinicalizing)를 동시에 제시한다. Interfacing(10/16) → zero-shot de novo design validation 한계 + AF3 adversarial limits + therapeutic protein design 체계 Clinicalizing(10/17).

실무 함의: SBDD 워크플로우팀에서 Garcia et al.의 zero-shot discrimination 결과를 de novo design screening criteria에 반영해야 한다.

*출처: [Garcia AlphaFold Zero-Shot De Novo Design Protein Science 2026](https://consensus.app/papers/details/f9c8c442c8065c0688de2538546747db/), [Feldman AlphaFold3 Adversarial Computational Structural Biotechnology 2026](https://consensus.app/papers/details/68f35194db525fb68a7bc8c10546001b/), [Mi Protein Design AI Era Computational Structural Biotechnology 2026](https://consensus.app/papers/details/5718c41069ab5c4da26a016fb3dd39ca/)*

*#ProteinAI #AlphaFold #Garcia #ProteinScience #ZeroShot #DeNovo #ProteinDesign #Discrimination #ESMFold #pLDDT #Topology #LogisticRegression #ConfidenceMetrics #Feldman #ComputationalStructuralBiotechnology #AlphaFold3 #Adversarial #40PercentInvariance #ConfidenceMetric #TemplateMemorization #Mi #ProteinDesign #AI #Transformer #TherapeuticProtein #Enzyme #SyntheticBiology #ClinicalizingDay #DeNovoDesign #Validation #BiomedicalAI #DailyReport*

---

### Post 5: Single-Cell Multi-Omics Foundation Model — Perturbation Benchmark 13 Methods × No Cross Cell-Type: Clinicalizing

🬬 **[Single-Cell Multi-Omics Foundation Model] Li et al.로 perturbation benchmark: 13 methods 25 datasets + PerturbNet superior DE + no method cross cell-type + conservative bias + foundation models compress variance(Science Advances 2026) + Yan et al.로 scGPT/Geneformer learnability: scaling doesn't guarantee improvement + perturbation prediction limited gains + real vs synthetic data( BMC Genomics 2026) + Dimitrov et al.로 single-cell interpretation review: causal inference + mechanistic discovery + disentanglement + population tracing + unifying ontology(Nature Reviews Genetics 2026): foundation model Clinicalizing(임상실현) 동시 확인**

Single-Cell Foundation Model의 **perturbation benchmark 체계화 + 한계 확인(Clinicalizing)**이 출처에서 동시에 확인된다[10][11][12].

**Perturbation Benchmark: PerturbNet Superior + No Cross Cell-Type Generalization[10]**
Lan-Xiang Li et al.(Science Advances 2026)의 연구는 13개 perturbation prediction 방법의 대규모 benchmark를 보고한다[10]. **13 representative methods 25 datasets**: perturbation modality/diversity species. **3 core tasks**: unseen single-gene generalization + combinatorial interactions + cross cell-type transfer. **24 metrics evaluation**: expression-level accuracy + delta + differential expression(DE) recovery + distributional similarity. **PerturbNet superior DE signature recovery Tasks 1 and 2**: PerturbNet DE 시그니처 회복 최고. **No method consistently generalized across cell types in Task 3**: cross cell-type 일반화 불발. **Models shared conservative bias**: 모든 모델 conservative bias 공유. **Foundation models compress variance and underestimate synergistic effects**: foundation model의 variance压缩 + 시너지 과소평가. **Different metrics probe distinct biological signals**: metrics가 서로 다른 biological signals 포착. 핵심: Perturbation benchmark — 이행의 방법론 한계(Clinicalizing).

**scGPT/Geneformer Learnability: Scaling Doesn't Guarantee Improvement[11]**
Yuhai Yan et al.(BMC Genomics 2026)의 연구는 single-cell foundation model의 learnability를 분석한다[11]. **Task-dependent benefits of large-scale pretraining**: task 의존적. **Substantial advantages in cell type annotation**: 세포 유형 주석에서 큰 이점. **Limited gains in perturbation prediction**:扰动 예측에서 제한적 이점. **Increasing model size doesn't guarantee improved performance**: bigger is better paradigma 도전. **May capture little more than simple summary statistics**:扰动 예측에서 단순 요약 통계 이상 포착 어려움. **Task-specific architectures and biologically-informed priors 필요성**: 향후 방향. 핵심: scGPT/Geneformer learnability — 이행의 scaling 한계(Clinicalizing).

**Single-Cell Interpretation Review: Unifying Ontology for Perturbation[12]**
Daniel Dimitrov et al.(Nature Reviews Genetics 2026)의 review는 single-cell 해석 방법론의 통합 온톨로지를 제시한다[12]. **Representation learning**: 표현 학습. **Causal inference**: 인과 추론. **Mechanistic discovery**: 기계적 발견. **Disentanglement**: 분리 학습. **Population tracing**: 집단 추적. **Unifying ontology**: 통합 분류 체계. **Online resource 제공**: companion resource. **Future computational directions**: 향후 전산 방향. 핵심: Single-cell interpretation — 이행의 방법론 통합 체계(Clinicalizing).

**10월 16일 Interfacing(ElSheikh Cell-JEPA + Whalley geometric outliers + Kendiukhov topology) vs 오늘의 차별점**: 10/16이 **ElSheikh ArXiv**으로 Cell-JEPA 36% improvement를, **Whalley bioRxiv**으로 scGPT/Geneformer/scFoundation geometric outliers를, **Kendiukhov PLOS One**으로 topology 0.80 CCA를 제시했다면, 오늘은 **Li Science Advances**으로 **13 methods 25 datasets benchmark + PerturbNet superior DE + no method cross cell-type + conservative bias + foundation models compress variance**라는 perturbation prediction의 체계적 한계(Clinicalizing)를, **Yan BMC Genomics**으로 **scaling doesn't guarantee improvement + perturbation prediction limited gains + real vs synthetic**라는 learnability 한계(Clinicalizing)를, **Dimitrov Nature Reviews Genetics**으로 **causal inference + mechanistic + disentanglement + population tracing + unifying ontology**라는 interpretation 방법론 체계(Clinicalizing)을 동시에 제시한다. Interfacing(10/16) → perturbation benchmark 한계 + scaling 한계 + interpretation ontology Clinicalizing(10/17).

실무 함의: 단일세포 연구팀에서 Li et al.의 perturbation benchmark 결과를 method selection criteria에 반영해야 한다.

*출처: [Li Perturbation Response Benchmark Science Advances 2026](https://consensus.app/papers/details/c9942dcafc665fd59f98c510a131e7e8/), [Yan scGPT Geneformer Learnability BMC Genomics 2026](https://consensus.app/papers/details/db23619e10075c0eb0957368f25ba7de/), [Dimitrov Single-Cell Interpretation Nature Reviews Genetics 2026](https://consensus.app/papers/details/4c9a5a9396a85b9797d6e1f0ecd4286b/)*

*#SingleCell #FoundationModel #Li #ScienceAdvances #Perturbation #Benchmark #13Methods #25Datasets #PerturbNet #DifferentialExpression #NoCrossCellType #ConservativeBias #VarianceCompression #Yan #BMCGenomics #scGPT #Geneformer #Learnability #Scaling #LimitedGains #Dimitrov #NatureReviewsGenetics #CausalInference #Mechanistic #Disentanglement #PopulationTracing #UnifyingOntology #ClinicalizingDay #PerturbationPrediction #MethodBenchmark #BiomedicalAI #DailyReport*

---

### Post 6: FDA/규제 — Kresladi 2026 Approval + 38 CGT Regulatory Assessment: Clinicalizing

📋 **[FDA/규제] Sheikh et al.로 Kresladi FDA approved March 2026: marnetegra-gene autotemcel + LAD-I + ITGB2 + March 26 2026(Annals Medicine Surgery 2026) + Oo et al.로 FDA 38 CGT approvals: 86.8% orphan + 92.1% expedited + 73.7% single study + 44.7% surrogate + postmarketing 73.4%(J Clinical Pharmacology 2026) + Shahzad et al.로 FDA CGT 38 approvals regulatory assessment: 1.6 mean pivotal studies + 17.9% primary efficacy postmarketing(Clinical Pharmacology Therapeutics 2026): FDA/규제 Clinicalizing(임상실현) 동시 확인**

FDA 규제 영역의 **Kresladi 2026 approval + 38 CGT regulatory assessment(Clinicalizing)**이 출처에서 동시에 확인된다[3][13][14].

**Kresladi FDA Approved March 2026 재확인[3]**
(M. Sheikh et al. Annals of Medicine and Surgery 2026 — Post 2에서 상세 기술). March 26, 2026 approval + LAD-I + ITGB2. 핵심: Kresladi FDA approval — 2026년 CGT 임상실현(Clinicalizing).

**FDA 38 CGT Approvals: 86.8% Orphan + 92.1% Expedited[13]**
C. Oo et al.(J Clinical Pharmacology 2026)의 연구는 FDA CGT 승인의 현황을 분석한다[13]. **38 CGTs approved through December 31, 2025**: FDA CGT 총 38개 승인. **86.8% orphan designation**: 희귀의약품 지위 높음. **92.1% expedited pathway**: 가속화 경로 활용 높음. **73.7% approved based on single study**: 단일 연구 기반 승인 다수. **44.7% relied solely on surrogate endpoints**: 대리 종착 근거 다수. **73.4% approved with postmarketing requirements or commitments**: 승인 후 연구 요구. **AI/ML-enabled analytics**: dose selection + safety evaluation + durability prediction. 핵심: FDA 38 CGT approvals — 이행의 규제 현실(Clinicalizing).

**FDA CGT Regulatory Assessment: 1.6 Mean Pivotal Studies + 17.9% Primary Efficacy Postmarketing[14]**
Mahnum Shahzad et al.(Clinical Pharmacology & Therapeutics 2026)의 연구는 FDA CGT 승인의 규제적 평가를 제시한다[14]. **Mean number of pivotal studies 1.6 (SD 1.7)**: 평균 1.6개 pivotal 연구. **73.7% approved based on single study**: 단일 연구 기반 다수. **44.7% relied solely on surrogate endpoints**: 대리 종착 근거 의존. **73.4% approved with postmarketing requirements**: 승인 후 연구 요구. **17.9% had postmarketing study with primary clinical efficacy endpoint**: 1차 임상 효능 종착 연구는 17.9%에 불과. **Most studies remain ongoing**: 대부분의 승인 후 연구 진행 중. **Implication**: timely access 우선 + limited premarket data + postmarketing efficacy 연구 부족. 핵심: FDA CGT regulatory assessment — 이행의 규제 평가 격차(Clinicalizing).

**10월 16일 Interfacing(Feierman FDA framework + Zhao demographics + Roskoski kinase) vs 오늘의 차별점**: 10/16이 **Feierman AJHG**으로 FDA personalized prime editing framework를, **Zhao International Journal**으로 demographic disparities를, **Roskoski Pharmacological Research**으로 kinase inhibitors를 제시했다면, 오늘은 **Sheikh Annals**으로 **Kresladi March 2026 approval + LAD-I**이라는 2026년 regulatory approval(Clinicalizing)을, **Oo J Clinical Pharmacology**으로 **38 CGTs + 86.8% orphan + 92.1% expedited + 73.7% single study**의 규제 현황 체계(Clinicalizing)을, **Shahzad Clinical Pharmacology Therapeutics**으로 **1.6 mean pivotal studies + 17.9% primary efficacy postmarketing**이라는 규제 평가 격차(Clinicalizing)을 동시에 제시한다. Interfacing(10/16) → 2026 approval + 38 CGT regulatory reality + assessment gaps Clinicalizing(10/17).

실무 함의: Regulatory Affairs팀에서 Shahzad et al.의 regulatory assessment 결과를 postmarketing strategy에 반영해야 한다.

*출처: [Sheikh Kresladi LAD-I FDA Approval Annals Medicine Surgery 2026](https://consensus.app/papers/details/f2f27f7392e05d9693a11da0cb87b29d/), [Oo FDA CGT 38 Approvals J Clinical Pharmacology 2026](https://consensus.app/papers/details/181658e93c825ea581a0664845a98d12/), [Shahzad FDA CGT Regulatory Assessment Clinical Pharmacology Therapeutics 2026](https://consensus.app/papers/details/24705d5037e85bcb9d9863f1aabca040/)*

*#FDA #GeneTherapy #Sheikh #Kresladi #FDA #Approved #March2026 #LADI #ITGB2 #Marnetegrave #Autotemcel #Oo #JClinicalPharmacology #38CGT #Orphan #Expedited #SingleStudy #Surrogate #Postmarketing #AIAnalytics #Shahzad #ClinicalPharmacology #Therapeutics #RegulatoryAssessment #PivotalStudies #PrimaryEfficacy #TimelyAccess #ClinicalizingDay #2026Approval #RegulatoryReality #AssessmentGaps #BiomedicalAI #DailyReport*

---

### Post 7: Spatial Transcriptomics — Pan-Cancer Niche + cfDNA Deep Learning Validation: Clinicalizing

🗺️ **[Spatial Transcriptomics] Li et al.로 pan-cancer 56 LCPs 13 niches: 12 cancer types 373 samples + Niche_4 poor prognosis + Niche_11 better survival(Cell Reports Medicine 2026) + Zhang et al.로 spatial ecotypes cfDNA: 9 SEs + deep learning + melanoma near 100 validation + immunotherapy response(Nature 2026) + Nguyen et al.로 spatial omics clinical TME: AI/ML + standardized workflows + precision medicine(Clinical & Translational Immunology 2026): spatial transcriptomics Clinicalizing(임상실현) 동시 확인**

Spatial transcriptomics의 **pan-cancer niche clinical outcomes + cfDNA validation(Clinicalizing)**이 출처에서 동시에 확인된다[15][16][17].

**Pan-Cancer 56 LCPs + 13 Niches: Niche_4/11 Survival Differential 재확인[15]**
(Jiarong Li et al. Cell Reports Medicine 2026 — 10/16에서 상세 기술). 12 cancer types + 373 samples + Niche_4(macrophage + tumor) poor prognosis + Niche_11(macrophage + immune) better survival. 핵심: Pan-cancer niche — 이행의 biomarker 체계(Clinicalizing).

**Spatial Ecotypes cfDNA Deep Learning: 9 SEs + Melanoma Near 100 Validation 재확인[16]**
(Wubing Zhang et al. Nature 2026 — 10/16에서 상세 기술). 9 conserved SEs + cfDNA deep learning + melanoma near 100명 validation + immunotherapy response. 핵심: Spatial ecotypes — 이행의 liquid biopsy validation(Clinicalizing).

**Spatial Omics Clinical TME: AI/ML + Standardized Workflows 재확인[17]**
(Hao Nguyen et al. Clinical & Translational Immunology 2026 — 10/16에서 상세 기술). AI/ML cross-platform integration + standardized workflows + precision cancer medicine. 핵심: Spatial omics clinical TME — 이행의 표준화 워크플로우(Clinicalizing).

**10월 16일 Interfacing(Chen MERFISH 2.0 + Wang spatial multi-omics + Li Niche_4/11) vs 오늘의 차별점**: 10/16이 **Chen Cancer Research**으로 MERFISH 2.0 1,000 genes + frozen/FFPE + custom 100 genes를, **Wang Cancer Research**으로 30 proteins × 1,000 RNA spatial multi-omics를, **Li 재확인**으로 Niche_4/11를 제시했다면, 오늘은 **Li/Zhang/Nguyen 모두 재확인**으로 **pan-cancer niche clinical outcomes + cfDNA deep learning validation + standardized workflows의 3개 모두**를 임상 실현(Clinicalizing) 관점에서 재해석한다. 오늘의 차별점: **재확인 출처들의 구체적 임상 실현 데이터**에 집중 — Li의 **Niche_4/11가 12개 암종 373개 검체에서 survival differential을 입증**했고(Zhang의 cfDNA validation과 함께 liquid biopsy prospective validation 가능성 제시), Zhang의 **near 100명 melanoma cfDNA deep learning validation**으로 **liquid biopsy 규모화 임상 실현 가능성**이 입증되었고, Nguyen의 **AI/ML standardized workflows**로 **clinical integration 체계화**가 제시되었다. Interfacing(10/16) → pan-cancer niche clinical outcomes + cfDNA validation + clinical workflows Clinicalizing(10/17).

실무 함의: 면역종양학팀에서 Zhang et al.의 cfDNA deep learning validation을 immunotherapy response prediction protocol에 반영해야 한다.

*출처: [Li Pan-Cancer Niche Cell Reports Medicine 2026](https://consensus.app/papers/details/f420ab985e1257e18352cac9e0c425cf/), [Zhang Spatial Ecotypes cfDNA Nature 2026](https://consensus.app/papers/details/b1d6ae0b47be530d8298e2eb72eff187/), [Nguyen Spatial Omics TME Clinical Translational Immunology 2026](https://consensus.app/papers/details/1b906e5abaaf5e789aaabde458127265/)*

*#SpatialTranscriptomics #Li #CellReportsMedicine #PanCancer #56LCPs #13Niches #Niche4 #Niche11 #Macrophage #Tumor #Prognosis #Survival #Zhang #Nature #SpatialEcotypes #cfDNA #DeepLearning #Melanoma #100Patients #Immunotherapy #Response #Nguyen #ClinicalTranslationalImmunology #SpatialOmics #TME #AIML #StandardizedWorkflows #PrecisionMedicine #ClinicalizingDay #ClinicalOutcomes #LiquidBiopsy #Validation #BiomedicalAI #DailyReport*

---

### Post 8: 오늘의 요약 — Clinicalizing(임상실현) Day 핵심 정리

🧬 **오늘 6개 영역 핵심 요약:**

10/17(Clinicalizing)의 핵심: Interfacing(10/16)된 Framework들이 **각 분야主流 체계에서 실제 임상 결과·실제 regulatory approval ·실제 experimental validation으로 이행(Clinicalizing)되는 날**이었다.

**1. Lonvoguran Ziclumeran Phase 3 RCT × Kresladi March 2026 Approval × SCD 148 Patients(Cohn NEJM 2026 + Sheikh Annals 2026 + Shaik European Journal 2026)**: Lonvo-z HAE Phase 3 -87% attack rate (P<0.001) + no serious/grade3 adverse events + NEJM publication; Kresladi March 26, 2026 FDA approval + LAD-I + ITGB2; exa-cel VF12 96.7% + lovo-cel VOE 100% + delivery 44/46 vs 35/47 — 임상 결과 + regulatory approval + delivery outcomes[1][2][3].

**2. AI Drug Discovery Phase IIa/Phase I Results + Multimodal/Digital Twins(Mao Frontiers Pharmacology 2026 + Seyhan Critical Reviews 2026 + Cai Pharmaceuticals 2026)**: Phase IIa encouraging efficacy + Phase I safety signals stalled + multimodal AI + digital twins + explainable AI — 임상 결과의 이중성 + preclinical pipeline 체계[4][5][6].

**3. AlphaFold Zero-Shot De Novo Design Validation + ProteinMPNN + Therapeutic Design(Garcia Protein Science 2026 + Feldman Computational Structural Biotechnology 2026 + Mi Computational Structural Biotechnology 2026)**: 614 benchmark monomers moderate discrimination + ESMFold pLDDT best + topology-dependent + failed designs better metrics; AF3 40% invariance + confidence failures; therapeutic protein + enzyme + synthetic biology — de novo design validation 한계 + therapeutic design 체계[7][8][9].

**4. Perturbation Benchmark 13 Methods × No Cross Cell-Type Generalization(Li Science Advances 2026 + Yan BMC Genomics 2026 + Dimitrov Nature Reviews Genetics 2026)**: PerturbNet superior DE + no method cross cell-type + conservative bias + foundation models compress variance; scaling doesn't guarantee improvement; causal inference + mechanistic + disentanglement + unifying ontology — perturbation prediction 방법론 한계 + 해석 체계[10][11][12].

**5. Kresladi 2026 Approval + 38 CGT Regulatory Assessment + 17.9% Primary Efficacy(Sheikh Annals 2026 + Oo J Clinical Pharmacology 2026 + Shahzad Clinical Pharmacology Therapeutics 2026)**: Kresladi March 2026 approval + LAD-I; 38 CGTs 86.8% orphan + 92.1% expedited + 73.7% single study + 44.7% surrogate; 1.6 mean pivotal studies + 17.9% primary efficacy postmarketing — 2026 regulatory approval + regulatory assessment gaps[3][13][14].

**6. Pan-Cancer Niche Clinical Outcomes + cfDNA Deep Learning Validation + Standardized Workflows(Li Cell Reports Medicine 2026 + Zhang Nature 2026 + Nguyen Clinical & Translational Immunology 2026)**: Niche_4/11 survival differential 12 cancer types 373 samples; 9 SEs cfDNA deep learning melanoma near 100 validation; AI/ML standardized workflows — clinical outcomes validation + liquid biopsy prospective[15][16][17].

10월 18일 전망: Clinicalizing된 Framework들이 다음 단계에서 어떻게 실제 환자 치료에 통합(Integrating)되는지가 핵심. 특히 Lonvoguran의 CRISPR Phase 3 성공이 in vivo CRISPR 질환 적용 확대, Kresladi approval이 rare disease gene therapy 규제 가속화, perturbation benchmark의 conservative bias 극복 방법이 관전이다.

Clinicalizing 다음은 **Integrating(통합)**이다. 오늘의 6개 전선이 제시하는 공통 질문: 임상 결과(Clinicalizing)된 Framework들이 각 분야主流 체계에서 어떤 방식으로 통합(Integrating)되어 실질적 환자 치료에 기여하는가. 이 질문이 내일 이후 각 전선의 핵심 과제가 된다.

*출처: [Shaik SCD Gene Therapy European Journal Haematology 2026](https://consensus.app/papers/details/f71effec863b55a39d2a9333578ff946/), [Cohn Lonvoguran Ziclumeran NEJM 2026](https://consensus.app/papers/details/014f88be1af55665a357eb28a20d5cbc/), [Sheikh Kresladi LAD-I FDA Approval Annals Medicine Surgery 2026](https://consensus.app/papers/details/f2f27f7392e05d9693a11da0cb87b29d/), [Mao AI Drug Discovery Frontiers Pharmacology 2026](https://consensus.app/papers/details/ca5321af0bb656e095b6108766231764/), [Seyhan AI Drug Discovery Critical Reviews Oncology Hematology 2026](https://consensus.app/papers/details/0e6b685fd2be56fb8a814115873a73d5/), [Cai Preclinical AI Pipeline Pharmaceuticals 2026](https://consensus.app/papers/details/9db7efc33cb250c4b08b448e73fcdd17/), [Garcia AlphaFold Zero-Shot De Novo Design Protein Science 2026](https://consensus.app/papers/details/f9c8c442c8065c0688de2538546747db/), [Feldman AlphaFold3 Adversarial Computational Structural Biotechnology 2026](https://consensus.app/papers/details/68f35194db525fb68a7bc8c10546001b/), [Mi Protein Design AI Era Computational Structural Biotechnology 2026](https://consensus.app/papers/details/5718c41069ab5c4da26a016fb3dd39ca/), [Li Perturbation Response Benchmark Science Advances 2026](https://consensus.app/papers/details/c9942dcafc665fd59f98c510a131e7e8/), [Yan scGPT Geneformer Learnability BMC Genomics 2026](https://consensus.app/papers/details/db23619e10075c0eb0957368f25ba7de/), [Dimitrov Single-Cell Interpretation Nature Reviews Genetics 2026](https://consensus.app/papers/details/4c9a5a9396a85b9797d6e1f0ecd4286b/), [Oo FDA CGT 38 Approvals J Clinical Pharmacology 2026](https://consensus.app/papers/details/181658e93c825ea581a0664845a98d12/), [Shahzad FDA CGT Regulatory Assessment Clinical Pharmacology Therapeutics 2026](https://consensus.app/papers/details/24705d5037e85bcb9d9863f1aabca040/), [Li Pan-Cancer Niche Cell Reports Medicine 2026](https://consensus.app/papers/details/f420ab985e1257e18352cac9e0c425cf/), [Zhang Spatial Ecotypes cfDNA Nature 2026](https://consensus.app/papers/details/b1d6ae0b47be530d8298e2eb72eff187/), [Nguyen Spatial Omics TME Clinical Translational Immunology 2026](https://consensus.app/papers/details/1b906e5abaaf5e789aaabde458127265/)*

*#ClinicalizingDay #Summary #Cohn #NEJM #Lonvoguran #Ziclumeran #HAE #Phase3 #87Percent #PValue #Sheikh #Kresladi #FDA #March2026 #LADI #Shaik #EuropeanJournal #SCD #148Patients #VF12 #VOE #DeliveryOutcomes #Mao #FrontiersPharmacology #AIDrugDiscovery #PhaseIIa #PhaseI #Safety #Multimodal #DigitalTwins #Garcia #ProteinScience #AlphaFold #ZeroShot #DeNovo #Discrimination #ESMFold #pLDDT #Li #ScienceAdvances #Perturbation #PerturbNet #NoCrossCellType #ConservativeBias #Yan #BMCGenomics #Scaling #Dimitrov #NatureReviews #Genetics #CausalInference #Mechanistic #Oo #JClinicalPharmacology #38CGT #Orphan #Expedited #Shahzad #ClinicalPharmacology #Therapeutics #PrimaryEfficacy #Postmarketing #Li #PanCancer #Niche4 #Niche11 #Zhang #cfDNA #DeepLearning #Melanoma #Nguyen #SpatialOmics #TME #Clinicalizing #ClinicalResults #RegulatoryApproval #Validation #BiomedicalAI #DailyReport*

---

## Sources (1차 출처 17개 이상)

[1] [Shaik M, et al. Efficacy, Safety, and Treatment-Delivery Feasibility of Autologous Gene Therapy for Sickle Cell Disease: A Systematic Review. European Journal of Haematology. 2026.](https://consensus.app/papers/details/f71effec863b55a39d2a9333578ff946/)
[2] [Cohn D, et al. Lonvoguran Ziclumeran for Hereditary Angioedema. New England Journal of Medicine. 2026.](https://consensus.app/papers/details/014f88be1af55665a357eb28a20d5cbc/)
[3] [Sheikh M, et al. Gene therapy for leukocyte adhesion deficiency type I: FDA approval, clinical evidence, and future perspectives. Annals of Medicine and Surgery. 2026.](https://consensus.app/papers/details/f2f27f7392e05d9693a11da0cb87b29d/)
[4] [Mao Z, et al. Artificial intelligence in drug discovery: from algorithmic foundations to clinical translation. Frontiers in Pharmacology. 2026.](https://consensus.app/papers/details/ca5321af0bb656e095b6108766231764/)
[5] [Seyhan A, et al. AI in drug discovery and development. Critical Reviews in Oncology/Hematology. 2026.](https://consensus.app/papers/details/0e6b685fd2be56fb8a814115873a73d5/)
[6] [Cai M, et al. From Algorithms to Assets: A Comprehensive Review of AI's Role in Preclinical Drug Discovery and the Hurdles to Clinical Translation. Pharmaceuticals. 2026.](https://consensus.app/papers/details/9db7efc33cb250c4b08b448e73fcdd17/)
[7] [Garcia M, et al. Evaluating zero-shot prediction of monomeric protein design success by AlphaFold, ESMFold, and ProteinMPNN. Protein Science. 2026.](https://consensus.app/papers/details/f9c8c442c8065c0688de2538546747db/)
[8] [Feldman J, et al. Adversarial Sequence Mutations in AlphaFold and ESMFold Reveal Nonphysical Structural Invariance, Confidence Failures, and Concerns for Protein Design. Computational and Structural Biotechnology Journal. 2026.](https://consensus.app/papers/details/68f35194db525fb68a7bc8c10546001b/)
[9] [Mi YL, et al. Protein Design Enters the Artificial Intelligence Era: Foundations, Tools, and Emerging Paradigms. Computational and Structural Biotechnology Journal. 2026.](https://consensus.app/papers/details/5718c41069ab5c4da26a016fb3dd39ca/)
[10] [Li LX, et al. A systematic comparison of single-cell perturbation response prediction models. Science Advances. 2026.](https://consensus.app/papers/details/c9942dcafc665fd59f98c510a131e7e8/)
[11] [Yan Y, et al. Evaluating the learnability of single-cell large language models on multiple tasks. BMC Genomics. 2026.](https://consensus.app/papers/details/db23619e10075c0eb0957368f25ba7de/)
[12] [Dimitrov D, et al. Interpretation, extrapolation and perturbation of single cells. Nature Reviews Genetics. 2026.](https://consensus.app/papers/details/4c9a5a9396a85b9797d6e1f0ecd4286b/)
[13] [Oo C, et al. FDA Gene Therapy Approvals (1998–2025): Current Status, Regulatory Evolution, and Future Directions. J Clinical Pharmacology. 2026.](https://consensus.app/papers/details/181658e93c825ea581a0664845a98d12/)
[14] [Shahzad M, et al. Assessment of United States Food and Drug Administration Approval of Cell and Gene Therapies. Clinical Pharmacology & Therapeutics. 2026.](https://consensus.app/papers/details/24705d5037e85bcb9d9863f1aabca040/)
[15] [Li J, et al. Pan-cancer analysis of spatial transcriptomics reveals heterogeneous tumor spatial microenvironment. Cell Reports Medicine. 2026.](https://consensus.app/papers/details/f420ab985e1257e18352cac9e0c425cf/)
[16] [Zhang W, et al. Non-invasive profiling of the tumour microenvironment with spatial ecotypes. Nature. 2026.](https://consensus.app/papers/details/b1d6ae0b47be530d8298e2eb72eff187/)
[17] [Nguyen H, et al. Spatial omics for profiling the dynamic tumor microenvironment. Clinical & Translational Immunology. 2026.](https://consensus.app/papers/details/1b906e5abaaf5e789aaabde458127265/)