NbFold enables accurate and rapid nanobody structure prediction through data-centric deep learning approach

NbFold is a deep learning framework for accurate and rapid nanobody structure prediction. This repository contains the trained model weights, benchmark datasets, prediction outputs, and evaluation results used in the accompanying paper. Contents:- Pretrained NbFold model weights- Benchmark nanobody datasets- Predicted structures- RMSD evaluation CSV files The model utilizes protein language model embeddings and structural refinement techniques for improved nanobody Fv prediction accuracy. These resources are provided to support reproducibility and further research in antibody and nanobody structure prediction.NbFold是一款可实现纳米抗体（nanobody）结构精准快速预测的深度学习框架。本仓库收录了配套论文中使用的已训练模型权重、基准数据集、预测输出结果与评估结果。 内容列表： - 预训练NbFold模型权重 - 基准纳米抗体数据集 - 预测得到的纳米抗体结构 - 均方根偏差（RMSD）评估CSV文件 该模型借助蛋白质语言模型（protein language model）嵌入表征与结构精修技术，提升了纳米抗体Fv区域的预测精度。本仓库提供上述全部资源，旨在支撑抗体与纳米抗体结构预测领域的可复现性研究及后续科研探索。