Accurate antibody affinity prediction is crucial for drug discovery. However, it remains challenging due to data scarcity, which limits the performance of deep learning models in few-shot, antigen-specific scenarios. To address this challenge, we propose MetaPrior, a novel meta-learning framework designed to learn robust prior knowledge before fine-tuning. The core of MetaPrior is a teacher-student mechanism where a Meta Network (teacher) is trained via bilevel optimization to generate high-quality labels for a diverse set of pseudo-samples and a binding affinity predictor (student) learns from these labels. The teacher is optimized by the student's performance on a small and experimentally measured guide set, enabling the pseudo-samples to approximate the true data distribution under conditions of data scarcity. Extensive experiments demonstrate that MetaPrior is compatible with popular network architectures (e.g., MLPs, CNN s and Transformers) and significantly surpasses standard fine-tuning strategies, particularly in data-scarce scenarios. Across several antigen-specific datasets, such as VEGF, MetaPrior achieves a notable performance gain even with only 40% of the available training data, increasing the Pearson correlation coefficient from 0.38 to 0.58 (a 53% improvement) and enhancing training stability by reducing the standard deviation from 0.42 to 0.19 (a 54.8% reduction in variability). These results confirm the robustness and generalizability of our proposed method. By jointly optimizing pseudo-labels and affinity learning, MetaPrior aligns synthetic supervision with the true data distribution, offering a principled solution for few-shot antibody affinity prediction.

Retinal diseases are leading causes of blindness globally. In real-world clinical practice, a patient may suffer from multiple retinal diseases, and these diseases are often under a long-tailed distribution, which poses significant challenges for accurate diagnosis. In this work, we propose a novel contrastive learning(CL)-based framework for multi-label retinal disease recognition. It consists of two parallel branches, a multi-label supervised contrastive learning branch and a classifier branch. The positive sets are created by the extent of proportional label overlap between samples and the anchor in calculating contrastive loss. For minority information enhancement, we design a hybrid-proxy model to generate class-dependent proxies, which are updated alongside the network. We capture rich relations samples, proxies, and labels by introducing the hybrid sample-proxy contrastive loss. Taking both label co-occurrence and data imbalance into consideration, we further utilize Distribution Balanced (DB) binary cross-entropy loss to guide the classifier branch learning. Experimental results on four public retinal disease datasets have demonstrated the superiority and effectiveness of our method.