AbLWR:A Context-Aware Listwise Ranking Framework for Antibody-Antigen Binding Affinity Prediction via Positive-Unlabeled Learning

Fan Xu; Zhi-an Huang; Haohuai He; Yidong Song; Wei Liu; Dongxu Zhang; Yao Hu; Kay Chen Tan

cs.LGApr 13, 2026

AbLWR:A Context-Aware Listwise Ranking Framework for Antibody-Antigen Binding Affinity Prediction via Positive-Unlabeled Learning

Authors: Fan Xu, Zhi-an Huang, Haohuai He, Yidong Song, Wei Liu, Dongxu Zhang, Yao Hu, Kay Chen Tan

Organizations: Department of Data Science and Artificial Intelligence, The Hong Kong Polytechnic University, Hong Kong SAR, China · Department of Computer Science, City University of Hong Kong (Dongguan), Dongguan, China · School of Computer Science and Engineering, Sun Yat-sen University, Guangzhou, Guangdong, China · School of Public Health, Xiamen University, Xiamen, China

Abstract

Accurate prediction of antibody-antigen binding affinity is fundamental to therapeutic design, yet remains constrained by severe label sparsity and the complexity of antigenic variations. In this paper, we propose AbLWR (Antibody-antigen binding affinity List-Wise Ranking), a novel framework that reformulates the conventional affinity regression task as a listwise ranking problem. To mitigate label sparsity, AbLWR incorporates a PU (Positive-Unlabeled) learning mechanism leveraging a dual-level contrastive objective and meta-optimized label refinement to learn robust representations. Furthermore, we address antigenic variation by employing a homologous antigen sampling strategy where Multi-Head Self-Attention (MHSA) explicitly models inter-sample relationships within training lists to capture subtle affinity nuances. Extensive experiments demonstrate that AbLWR significantly outperforms state-of-the-art baselines, improving the Precision@1 (P@1) by over 10 $\%$ in randomized cross-validation experiments. Notably, case studies on Influenza and IL-33 validate its practical utility, demonstrating robust ranking consistency in distinguishing subtle viral mutations and efficiently prioritizing top-tier candidates for wet-lab screening.

Figures & tables

Figure 1: Key challenges in Ab-Ag binding affinity prediction. (a) Distinct affinity profiles of multiple antibodies binding to the same antigen. (b) The large reservoir of unlabeled data remains largely unexploited by existing supervised paradigms, compared with the immense volume of antibody and antigen sequences. (c) Analysis of existing binding affinity data from public datasets (refer to Appendix Section A.1 for data collection details) reveals that most antigens are associated with different antibody candidates, whereas individual antibodies typically bind to a single specific antigen.

Figure 2: The overall framework of AbLWR. The architecture comprises three key components: (a) PU Pre-training: Dual GNN encoders learn robust representations from the combined dataset ( $\mathcal{D}$ ) by jointly optimizing a composite contrastive objective ( $\mathcal{L}_{\text{Ins}}+\mathcal{L}_{\text{Clus}}$ ) and a cross entropy loss ( $\mathcal{L}_{\text{CE}}$ ) with meta-reweighted pseudo-labels. (b) List Sampling: A homologous sampling strategy constructs informative training lists $\mathcal{T}$ based on antigen sequence similarity ( $\delta_{\text{seq}}$ ) and affinity margins ( $y_{\text{cutoff}}$ ). (c) MHSA Ranking: The sampled lists are processed by MHSA mechanism to capture intra-list interactions via self-attention, yielding the final affinity ranking scores $\mathbf{r}$ .

Category	Model	FRA (%)	$K\tau$	PRA (%)	PAU (%)	P@1 (%)
Sequence-based	MINT	1.11 $\pm$ 0.09	0.07 $\pm$ 0.00	53.33 $\pm$ 0.16	53.33 $\pm$ 0.16	23.12 $\pm$ 0.62
Sequence-based	AntiBERTy	5.09 $\pm$ 0.19	0.35 $\pm$ 0.01	67.59 $\pm$ 0.43	67.59 $\pm$ 0.43	39.96 $\pm$ 2.45
Structure-based	AbRank	5.63 $\pm$ 2.78	0.31 $\pm$ 0.11	65.40 $\pm$ 5.61	65.40 $\pm$ 5.61	39.64 $\pm$ 5.65
Structure-based	GraphDTA	8.98 $\pm$ 1.23	0.46 $\pm$ 0.03	73.02 $\pm$ 1.68	73.02 $\pm$ 1.68	46.92 $\pm$ 4.90
Complex-based	ANTIPASTI ∗	0.00 $\pm$ 0.00	-0.07 $\pm$ 0.07	46.24 $\pm$ 3.69	46.24 $\pm$ 3.69	17.96 $\pm$ 4.42
Traditional	LambdaMART	6.36 $\pm$ 0.24	0.38 $\pm$ 0.01	68.18 $\pm$ 0.64	68.79 $\pm$ 0.63	39.04 $\pm$ 0.48

Table 1: Performance comparison on the Random Split. We employ 5-fold cross-validation (mean

\pm

std) for all methods. Due to computational costs, the methods of FoldX and ANTIPASTI (marked with ∗ ) were evaluated on a subsampled dataset (see Appendix Section A.2 for details). Accordingly, AbLWR ∗ reports our model’s performance on this specific subset to ensure a fair comparison. All metrics are in percentage (%) except for

K\tau

, and higher values indicate better performance. Best results are highlighted in bold .

Ag-based Split
Category	Model	FRA (%)	$K\tau$	PRA (%)	PAU (%)	P@1 (%)
Sequence- based	MINT	0.67	0.00	49.63	49.63	21.93
Sequence- based	AntiBERTy	3.33	0.31	64.98	64.98	34.50
Structure- based	AbRank	1.61	0.15	57.20	57.20	24.21
Structure- based	GraphDTA	6.02	0.42	70.96	70.96	41.59
Complex- based	ANTIPASTI ∗	0.00	-0.01	49.39	49.39	22.45

Table 2: Performance comparison on Ag-based and Ab-based Splits. The Ag-based split corresponds to the single most divergent cluster, while the Ab-based split comprises the two most distinct clusters.

Sampling Strategy
Model	FRA (%)	$K\tau$	PRA (%)	PAU (%)	P@1 (%)
AbLWR (Ours)	21.28	0.60	79.96	80.14	57.73
w/ AbRankSampling	12.55	0.52	75.73	75.90	52.92
Component Ablation
w/o PU Init.	5.52	0.34	66.08	66.25	41.67
w/o Rank. ListMLE	3.29	0.56	59.17	59.30	31.55

Table 3: Ablation study on the test set. We evaluate the impact of different components and strategies.

Figure 3: Visualization of ranking predictions on 18 sampled lists. The heatmap displays the ground truth ranks mapped to the model’s predicted positions. Each row represents a test list colored by their ground truth rank, and columns Pred1–Pred5 denote the predicted order. The adjacent bar chart shows the corresponding $K\tau$ calculated for each list row.

Figure 4: Case studies on FNI17 antibody mechanism and candidates screening for IL-33 antigen. (a) Comparison of ground truth ranks and predicted ranks. Lines connect the same strain across the two rankings. (b) Dual-axis plot showing the ground truth values and predicted scores for the corresponding viral strains. (c) Hit rate curve for identifying the top-1 binder to the IL-33 antigen. (d) Recall curve showing the percentage of top-3 IL-33 binders retrieved against the number of candidates screened.

Appendix figures & tables9 assets

Supplementary material from the paper’s appendix.

Appendix

Figure 5: Binding affinity statistics. (a) Affinity variance per antigen. (b) Affinity distribution for the top 10 most represented antigen groups.

Figure 6: Visualization of sequence clustering. (a) Antigen sequence clustering. Cluster 0 is identified as a divergent group distinct from the main distribution and is used to define the Ag-based split. (b) Antibody sequence clustering based on CDRs. Cluster 1 and Cluster 5 exhibit significant deviation from the main trend and are selected for the Ab-based split. In both scenarios, these distinct clusters serve as the primary source for the test sets to evaluate model generalization capabilities.

PU Pre-training		MHSA Ranking
Parameter	Value	Parameter	Value
Optimizer	SGD	Optimizer	Adam
Max Epochs	400	Max Epochs	50
Warm-up Epochs	20	Batch Size	16
Batch Size	64	Learning Rate	0.001
Learning Rate	0.001	Weight Decay	0.3

Appendix

Table 4: Hyperparameter settings for AbLWR. The model is trained in two stages: PU Pre-training (left) and MHSA Ranking (right).

			In vitro neutralization IC 50 ( $\mu$ g/mL)
			FNI9		FNI17		FNI19
Strain	Abbr.	Group	IC 50	PDB	IC 50	PDB	IC 50	PDB
A/Darwin/6/2021	A/DW/2021	H3N2	1.128	–	$>50$	–	2.790	–
A/HongKong/2671/2019	A/HK/2019	H3N2	0.073	8G3P	5.068	–	0.190	8G40
A/Singapore/INFIMH-16-0019/2016	A/SA/2016	H3N2	0.162	–	$>50$	–	0.916	–
B/Massachusetts/2/2012	B/MA/2012	Yamagata	0.204	–	0.125	–	0.472	8G3Z

Appendix

Table 5: In vitro neutralization IC 50 values (

\mu

g/mL) and corresponding PDB codes. Data are adapted from Momont et al. ( Momont et al., 2023 ) . Lower IC 50 values indicate higher potency. Entries marked as ‘

>50

’ highlight the loss of neutralization efficacy (immune escape). ‘–’ denotes that the crystal structure is unavailable.

Metric	Influenza Virus	Human IL-33
Kendall’s $\tau>0.2$	67.83%	45.32%
Kendall’s $\tau>0.4$	43.32%	27.43%
Kendall’s $\tau>0.6$	18.02%	13.26%

Appendix

Table 6: Quantitative evaluation of ranking fidelity for AbLWR on influenza virus and human IL-33 datasets. The values represent the percentage of test cases where the

K\tau

exceeds the specified thresholds (

0.2,0.4,0.6

Figure 7: Detailed visualization of ranking performance on (a) influenza viruses and (b) human IL-33. The comparison includes (left) Confusion Matrices between predicted and ground truth ranks, (middle) Distribution of Ground Truth Ranks for candidates predicted as Top-1, and (right) Boxplots showing the true affinity value distribution across predicted rank positions.

Figure 8: Visualization of ranking predictions on 18 sampled lists by four representative data driven baseline models. (a) AntiBERTy (sequence-based); (b) AbRank (structure-based pairwise ranking); (c) GraphDTA (structure-based); and (d) ANTIPASTI (complex-based).

Figure 9: Analysis of FNI17 antibody predictions by four baselines. Models: (a) AntiBERTy (sequence-based); (b) AbRank (structure-based pairwise ranking); (c) GraphDTA (structure-based); (d) ANTIPASTI (complex-based).

Cand.	$\text{EC}_{50}$	Cand.	$\text{EC}_{50}$	Cand.	$\text{EC}_{50}$	Cand.	$\text{EC}_{50}$
I7-WT	11.41	I7-M5	1898	I7-M10	1.712	I7-M15	705.3
I7-M1	2.561	I7-M6	462.2	I7-M11	4.16	I7-M16	1.402
I7-M2	37.15	I7-M7	2.898	I7-M12	16.2	I7-M17	1.326
I7-M3	148.1	I7-M8	5961	I7-M13	9.857	I7-M19	2.134
I7-M4	12.79	I7-M9	9.845	I7-M14	67.34	I7-M20	9.352

Appendix

Table 7: Experimental

\text{EC}_{50}

values (nM) of the I7 single-chain antibody and its variants against Human IL-33 ( Wang et al., 2025 ) . The dataset comprises the wild-type (I7-WT) and 19 generated variants, where lower values indicate stronger binding affinity.

Learning to Rank

Learning to rank develops models that order items by relevance or utility for search, recommendations, and other decisions.

See all 266 papers

AbICL: In-Context Learning for Antigen-Specific Antibody Affinity Ranking
Jul 7, 2026Zhiyuan Chen, Jing Hu, Junzhe Wang +4Learning to Rank Meta-Learning
Henlius Shanghai, China
DuaDeep-SeqAffinity: Dual-Branch Deep Learning for Tri-Stream Sequence-Based Antibody--Antigen Affinity Prediction
Dec 26, 2025Aicha Boutorh, Soumia Bouyahiaoui, Manel Kara Laouar +3Hybrid CNN-Transformer Architectures Molecular Property Prediction
National School of Artificial Intelligence (ENSIA) Sidi Abdallah Campus, Algiers, Algeria · Lamine Debaghine University Hospital Center, CHU Bab El Oued, Algiers, Algeria
AbAffinity: A Large Language Model for Predicting Antibody Binding Affinity against SARS-CoV-2
Mar 4, 2026Faisal Bin Ashraf, Animesh Ray, Stefano LonardiProtein Language Models
Department of Computer Science and Engineering,University of California, Riverside, 92521, CA, United States · Riggs School of Applied Life Sciences, Keck Graduate Institute, Claremont, 91711, CA, United States

Jul 7, 2026J/K move · Enter open · S save

AbICL: In-Context Learning for Antigen-Specific Antibody Affinity Ranking

Zhiyuan Chen, Jing Hu, Junzhe Wang, Yueyang Huang+3

Henlius Shanghai, China

Learning to Rank Meta-Learning In-Context Learning

Open paper AskPDF arXiv

Accurate ranking of antibody candidates according to their binding affinity is essential for therapeutic antibody discovery. However, existing methods treat affinity comparisons independently and ignore the contextual information encoded in other labeled comparisons, limiting their ability to capture antigen-specific binding landscapes. For many target antigens, a small number of experimentally characterized affinity comparisons are often available. An important question is whether the model can exploit these existing comparisons to infer antigen-specific ranking patterns that facilitate subsequent affinity ranking. This form of learning from labeled demonstrations closely resembles the paradigm of In-Context Learning, motivating us to revisit antibody affinity ranking from an ICL perspective. To this end, we propose AbICL, an ICL framework for antigen-specific antibody affinity ranking. AbICL combines a pretrained structural encoder with a context ranking head and is trained with an episodic meta-training strategy that enables the model to leverage support demonstrations for test-time adaptation without gradient updates. Experiments on the AbRank benchmark demonstrate that AbICL consistently outperforms existing ranking baselines across almost all data splits and evaluation benchmarks. Further analysis shows that the value of contextual demonstrations depends on how well they match the target inference task, and becomes increasingly pronounced under distribution shift and fine-grained affinity discrimination. These findings highlight the potential of ICL as an effective paradigm for antigen-specific antibody affinity ranking, particularly in challenging settings where a single global ranking function is insufficient.

Abstract

Learning to Rank

Explore similar work

AbICL: In-Context Learning for Antigen-Specific Antibody Affinity Ranking

DuaDeep-SeqAffinity: Dual-Branch Deep Learning for Tri-Stream Sequence-Based Antibody--Antigen Affinity Prediction

AbAffinity: A Large Language Model for Predicting Antibody Binding Affinity against SARS-CoV-2