Brief · 2026-10-21

Personalized neoantigen vaccines are moving from proof-of-concept into treatment-paradigm territory: Geneos enters the adjuvant setting in resected liver cancer with a randomized Phase 2, and Merck/Moderna's intismeran autogene heads to ESMO 2026 as the meeting's most anticipated dataset.PR Newswire BioSpace

On the enabling side, LG Chem laid out an AI-driven plan spanning neoantigen prediction and response-biomarker-based patient selection targeting 2029 clinical entry, while a new method (EMBER) reports more robust TCR-pMHC binding prediction by eliminating docked-structure error.The Korea Herald BigGo Finance arXiv The thread running through today's items: category momentum is shifting toward adjuvant and curative-intent strategies, and differentiation is migrating to patient selection.

ClinicaltodayNew

Geneos Therapeutics Initiates IMPACT31 Phase 2 Randomized Clinical Trial Evaluating Neoantigen-Targeted Personalized DNA Immunotherapy for Cancer - PR Newswire

Geneos Therapeutics initiated IMPACT31 (NCT07722741), a global multicenter Phase 2 randomized trial planning to enroll 90 patients with high-risk hepatocellular carcinoma who have undergone curative-intent resection, testing its personalized DNA immunotherapy GNOS-PV02 against active surveillance. The adjuvant design targets recurrence prevention in a post-surgical population, a setting with a cleaner regulatory path than advanced disease if it reads out. The company also appointed Julie Rosenberg, MD, a biopharma veteran with more than two decades of drug development experience, as Chief Medical Officer to drive the trial.

news · today · PR Newswire

Also new today04

CommercialtodayNew

LG Chem eyes 2029 trials for personalized cancer vaccine - The Korea Herald

LG Chem said it aims to begin clinical development of a personalized mRNA cancer vaccine in 2029, with preclinical testing planned for 2027, using AI to predict patient-specific neoantigens. The strategy, unveiled at the inaugural LG AI Bio Conference in Seoul co-hosted with LG AI Research and Flagship Pioneering, also includes AI models to identify biomarkers that predict treatment response and guide trial patient selection. Life sciences head Son Jee-woong framed the thesis as finding not just the right drug but the right patient - a long-dated entrant into a field where clinical-stage peers are already running randomized Phase 2 trials.

news · today · The Korea Herald

CommercialtodayNew

LG Chem Takes AI Into Clinical Patient Selection, Targets 2029 Trial for Personalized Cancer Vaccine - BigGo Finance

LG Chem (051910) is extending AI beyond drug candidate discovery into clinical patient selection and treatment strategy design for its personalized mRNA cancer vaccine program, targeting 2029 trial entry. The 2026 LG AI Bio Conference, held at LG Science Park in Seoul with LG AI Research and Flagship Pioneering, drew roughly 140 executives from domestic and international AI-bio companies. The differentiator being pitched is AI-driven discovery of response-predictive biomarkers from complex biological data - the same patient-selection capability likely to separate winners as personalized vaccines advance.

news · today · BigGo Finance

Clinical2 days agoNew

Merck/Moderna, Summit/Akeso and ADCs to headline ESMO 2026 - BioSpace

ESMO 2026 opens in Madrid with Merck and Moderna's mRNA-based personalized cancer vaccine intismeran autogene plus Keytruda as the most anticipated presentation; in August the partners reported the combination extended survival without disease recurrence. BioSpace also flags Summit Therapeutics/Akeso's PD-1xVEGF bispecific and antibody-drug conjugate data as headliners. Stifel's Dara Azar frames the investor question directly: after a cancer vaccine win, where is the next opportunity - a sign that this dataset could reset expectations across the personalized vaccine category.

news · 2 days ago · BioSpace

AI / Methods2 days agoNew

EMBER: Predicting TCR--pMHC Binding Without a Docked Complex

EMBER is a structure-based TCR-pMHC binding predictor that folds the TCR and pMHC independently and compares SE(3)-invariant residue-graph representations without constructing a shared coordinate frame, eliminating the docked-complex placement error that propagates through traditional predictors while cutting preprocessing from O(NM) to O(N+M). Under epitope-disjoint evaluation against six retrained baselines across four folding tools, it ranked first on every metric: gains of 0.9-14.4% in micro-AUROC, 18.5-25.3% in micro-AUPRC, and 10.6-24.1% in MCC, with rankings up to 85% more stable across folding tools. For vaccine developers, immunogenicity prediction is the upstream constraint on neoantigen selection, so folding-tool-agnostic accuracy directly bears on target-prioritization reliability.

arxiv · 2 days ago · Lin Ha, Chen Zhou, Haozheng Luo, Han Liu, Xiaoyu Zhang, Yan Chen

Geneos enters adjuvant liver cancer with randomized Phase 2