ODesign: A World Model for Biomolecular Interaction Design

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Zhang, Odin, Zhang, Xujun, Lin, Haitao, Tan, Cheng, Wang, Qinghan, Mo, Yuanle, Feng, Qiantai, Du, Gang, Yu, Yuntao, Jin, Zichang, You, Ziyi, Lin, Peicong, Zhang, Yijie, Tao, Yuyang, Chen, Shicheng, Chen, Jack Xiaoyu, Hua, Chenqing, Zhao, Weibo, Ma, Runze, Xia, Yunpeng, Ying, Kejun, Li, Jun, Zeng, Yundian, Lang, Lijun, Pan, Peichen, Cao, Hanqun, Song, Zihao, Qiang, Bo, Wang, Jiaqi, Ji, Pengfei, Bai, Lei, Zhang, Jian, Hsieh, Chang-yu, Heng, Pheng Ann, Sun, Siqi, Hou, Tingjun, Zheng, Shuangjia
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866909872866459648
author Zhang, Odin
Zhang, Xujun
Lin, Haitao
Tan, Cheng
Wang, Qinghan
Mo, Yuanle
Feng, Qiantai
Du, Gang
Yu, Yuntao
Jin, Zichang
You, Ziyi
Lin, Peicong
Zhang, Yijie
Tao, Yuyang
Chen, Shicheng
Chen, Jack Xiaoyu
Hua, Chenqing
Zhao, Weibo
Ma, Runze
Xia, Yunpeng
Ying, Kejun
Li, Jun
Zeng, Yundian
Lang, Lijun
Pan, Peichen
Cao, Hanqun
Song, Zihao
Qiang, Bo
Wang, Jiaqi
Ji, Pengfei
Bai, Lei
Zhang, Jian
Hsieh, Chang-yu
Heng, Pheng Ann
Sun, Siqi
Hou, Tingjun
Zheng, Shuangjia
author_facet Zhang, Odin
Zhang, Xujun
Lin, Haitao
Tan, Cheng
Wang, Qinghan
Mo, Yuanle
Feng, Qiantai
Du, Gang
Yu, Yuntao
Jin, Zichang
You, Ziyi
Lin, Peicong
Zhang, Yijie
Tao, Yuyang
Chen, Shicheng
Chen, Jack Xiaoyu
Hua, Chenqing
Zhao, Weibo
Ma, Runze
Xia, Yunpeng
Ying, Kejun
Li, Jun
Zeng, Yundian
Lang, Lijun
Pan, Peichen
Cao, Hanqun
Song, Zihao
Qiang, Bo
Wang, Jiaqi
Ji, Pengfei
Bai, Lei
Zhang, Jian
Hsieh, Chang-yu
Heng, Pheng Ann
Sun, Siqi
Hou, Tingjun
Zheng, Shuangjia
contents Biomolecular interactions underpin almost all biological processes, and their rational design is central to programming new biological functions. Generative AI models have emerged as powerful tools for molecular design, yet most remain specialized for individual molecular types and lack fine-grained control over interaction details. Here we present ODesign, an all-atom generative world model for all-to-all biomolecular interaction design. ODesign allows scientists to specify epitopes on arbitrary targets and generate diverse classes of binding partners with fine-grained control. Across entity-, token-, and atom-level benchmarks in the protein modality, ODesign demonstrates superior controllability and performance to modality-specific baselines. Extending beyond proteins, it generalizes to nucleic acid and small-molecule design, enabling interaction types such as protein-binding RNA/DNA and RNA/DNA-binding ligands that were previously inaccessible. By unifying multimodal biomolecular interactions within a single generative framework, ODesign moves toward a general-purpose molecular world model capable of programmable design. ODesign is available at https://odesign.lglab.ac.cn ,
format Preprint
id arxiv_https___arxiv_org_abs_2510_22304
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle ODesign: A World Model for Biomolecular Interaction Design
Zhang, Odin
Zhang, Xujun
Lin, Haitao
Tan, Cheng
Wang, Qinghan
Mo, Yuanle
Feng, Qiantai
Du, Gang
Yu, Yuntao
Jin, Zichang
You, Ziyi
Lin, Peicong
Zhang, Yijie
Tao, Yuyang
Chen, Shicheng
Chen, Jack Xiaoyu
Hua, Chenqing
Zhao, Weibo
Ma, Runze
Xia, Yunpeng
Ying, Kejun
Li, Jun
Zeng, Yundian
Lang, Lijun
Pan, Peichen
Cao, Hanqun
Song, Zihao
Qiang, Bo
Wang, Jiaqi
Ji, Pengfei
Bai, Lei
Zhang, Jian
Hsieh, Chang-yu
Heng, Pheng Ann
Sun, Siqi
Hou, Tingjun
Zheng, Shuangjia
Biomolecules
Biomolecular interactions underpin almost all biological processes, and their rational design is central to programming new biological functions. Generative AI models have emerged as powerful tools for molecular design, yet most remain specialized for individual molecular types and lack fine-grained control over interaction details. Here we present ODesign, an all-atom generative world model for all-to-all biomolecular interaction design. ODesign allows scientists to specify epitopes on arbitrary targets and generate diverse classes of binding partners with fine-grained control. Across entity-, token-, and atom-level benchmarks in the protein modality, ODesign demonstrates superior controllability and performance to modality-specific baselines. Extending beyond proteins, it generalizes to nucleic acid and small-molecule design, enabling interaction types such as protein-binding RNA/DNA and RNA/DNA-binding ligands that were previously inaccessible. By unifying multimodal biomolecular interactions within a single generative framework, ODesign moves toward a general-purpose molecular world model capable of programmable design. ODesign is available at https://odesign.lglab.ac.cn ,
title ODesign: A World Model for Biomolecular Interaction Design
topic Biomolecules
url https://arxiv.org/abs/2510.22304