Skip to content
Log in
#small language model Open access

Using multimodal foundational models to predict neoantigen immunogenicity and vaccine effectiveness across different tumor types

Gang Liu Jia Wang Jia Zhu
Aug 2026 · Global Health Care · 0 citations · 21 references

TL;DR

This work is the first to use a multi-modal foundation model for neoantigen vaccines, overcoming the tumor-type-specific constraints of conventional approaches and facilitating joint modelling of immunogenicity and clinical efficacy, thereby offering an AI decision engine for precision cancer vaccine design that is applicable to all cancer types.

Abstract

Neoantigen vaccines are a key component of personalised cancer immunotherapy; however, existing prediction techniques are primarily restricted to a single tumour type and have difficulty integrating multi-modal biological data, which leads to inadequate accuracy in immunogenicity evaluation and vaccine efficacy prediction. In order to predict neoantigen immunogenicity and customised vaccination clinical outcomes across various tumour types, this study attempts to develop and verify a universal multi-modal foundation model.Neoantigen peptide sequences, mass spectrometry-derived pHLA binding patterns, single-cell TCR repertoires, tumour transcriptomes, and clinical vaccination trial follow-up records were among the extensive paired data we gathered from 15 solid tumour types. We present the NeoVAX-FM multi-modal foundation model, which first embeds peptide sequences, pHLA complex 3D structures, and gene expression into a single semantic space using a contrastive language-image pretraining paradigm. It is then fine-tuned on downstream tasks to concurrently score immunogenicity and predict progression-free survival.The model was assessed in one prospective clinical trial and three external validation cohorts. NeoVAX-FM showed strong performance in melanoma, non-small cell lung cancer, and microsatellite stable colorectal cancer, with an average AUC of 0.94 for cross-tumor neoantigen immunogenicity prediction—a 12.3% increase over the best currently available techniques. Patients in the prospective vaccination cohort who were projected by the model to be “high responders” had a considerably higher median progression-free survival (HR = 0.28, p < 0.001), and the model was successful in identifying tumour microenvironment characteristics and universal TCR motifs that drive long-term responses.This work is the first to use a multi-modal foundation model for neoantigen vaccines, overcoming the tumor-type-specific constraints of conventional approaches and facilitating joint modelling of immunogenicity and clinical efficacy, thereby offering an AI decision engine for precision cancer vaccine design that is applicable to all cancer types.

DOI Semantic Scholar
Read PDF

Similar papers

#small language model Open access Aug 2026

LifeSciBench: Evaluating Language Models on Realistic, Expert-Level Tasks in the Life Sciences

LifeSciBench is introduced, a benchmark of 750 expert-authored tasks designed to evaluate whether language models can handle realistic life science research work, with each constituent task paired with a human expert-written rubric.

Amelia Liu, Andrew Ho, Anne Marie Droste et al. · 2 citations
#small language model Preprint Aug 2026

CritICL: Inference-Time Weak-to-Strong Generalization from Small Language Model Failure Modes

Experimental results show that CritICL consistently outperforms standard in-context learning and achieves performance competitive with or superior to test-time scaling methods, while requiring significantly fewer generations and lower token cost.

Yu-Fan Wu, Yinghui He, Zhengyi Hu et al. · 1 citation
#artificial intelligence Preprint Aug 2026

TestifAI: Tomography-Based Testing for Deep Learning Systems

TestifAI, a deep learning testing framework for efficient and accurate estimation of robustness against combinations of perturbations, is proposed and partial model tomography is introduced, a novel approach to reconstructing model behaviour in a multi-perturbation space from tests that apply only a small number of perturbations.

Arooj Arif, T. Hartung, E. Botoeva et al. · 1 citation
#small language model Sep 2026

Low Carbon Scheduling of Integrated Energy System Based on Large Language Model-Embedded Multi-Agent Reinforcement Learning

The complex multi-energy coupling characteristics inherent to integrated energy system (IES) present unprecedented challenges for the implementation of low-carbon scheduling. Existing optimization methods often exhibit limitations in system scalability, algorithm adaptivity, and carbon reduction efficacy for complex IES. This paper proposes a Large Language Model (LLM)-Embedded Multi-Agent Reinforcement Learning (LEMARL) to address the aforementioned issues. The proposed method integrates the global perception capability of LLMs with the dynamic optimization capability of MARL. Specifically, the LLM-Embedded module generates high-quality reward functions and policy frameworks from a global perspective, while the MARL module leverages these LLM-generated strategies for distributed interactive iterations—greatly enhancing computation efficiency and scalability. Simulation results demonstrate that LEMARL reduces carbon emissions by 7.76% and simultaneously decreases operating costs by 4.49% in a small-scale IES. Furthermore, LEMARL also exhibits superior applicability and scalability in large-scale IES of the IEEE 141-bus power grid integrated with 51-node thermal system.

Chen Xia, Tong Gou, Yinliang Xu et al. · 1 citation
#small language model Preprint Aug 2026

Apodex 1.1: Scaling Agentic Intelligence for Complex Work

General-purpose language models can reason and synthesize knowledge, but complex work also requires sustained interaction with files, information sources, and executable code, together with state maintenance, failure recovery, and verifiable delivery. We call this \emph{working capability}: sustained, verifiable progress toward a real-world objective. Apodex 1.1 develops this capability along two complementary dimensions. \emph{Environment Scaling} expands the diversity and verifiability of executable file, search, and code environments, while \emph{Agentic Coordination Scaling} trains agents to decompose long-horizon tasks, delegate parallel work, integrate asynchronous results, and replan. A shared execution harness and AgentOS maintain task state and provenance across tools and agents, and training turns environment trajectories and coordination traces into reliable behavior. Across complex professional work, finance, scientific research, mathematics, coding, and search, Apodex 1.1 reaches the leading performance band despite using a substantially smaller model than many frontier systems. The 35B-parameter Apodex 1.1 Mini further retains strong working capability in a locally deployable form. These results ground agentic intelligence in useful, verifiable work completed over time and advance our goal of building a \emph{Heavy-Duty Solver} for ambitious, long-running tasks.

Apodex Team B. An, B. Li, B. Wang et al. · 1 citation

Related blog posts