---
title: 'Ab-Affinity: SARS-CoV-2 Binding Predictor'
url: https://www.emergentmind.com/topics/ab-affinity
type: topic
---

# Ab-Affinity: SARS-CoV-2 Binding Predictor

Searching arXiv for the primary paper and closely related antibody-affinity works to ground the article.
Ab-Affinity is a sequence-based antibody affinity predictor specialized for SARS-CoV-2-related binding, introduced as a large language model for estimating how strongly a single-chain fragment variable (scFv) antibody variant binds a fixed target peptide from the viral spike protein, specifically a conserved peptide in the HR2 region [2603.04480]. In this formulation, the model does not encode an arbitrary paired antigen sequence; rather, the antigen is implicit in the task definition because all supervised measurements concern one SARS-CoV-2 HR2 peptide [2603.04480]. The system is therefore best understood as a target-specialized affinity regressor built by fine-tuning a pretrained protein language model so that antibody sequence representations become predictive not only of binding affinity itself, but also of affinity classes, improvement over a seed antibody, residue-level interaction patterns, and apparent thermostability trends [2603.04480].

## 1. Biological problem and task scope

Binding affinity is one of the central determinants of whether an antibody will be useful as a neutralizer or therapeutic lead [2603.04480]. Antibodies recognize antigens through paratopes, typically formed by complementarity-determining regions in the variable domains of the heavy and light chains, and strong and specific binding is often required for effective neutralization [2603.04480]. In practice, affinity measurement depends on assays such as Surface Plasmon Resonance, ELISA, or Bio-Layer Interferometry, together with generation and purification of many candidate antibodies, making the process expensive and slow [2603.04480]. This creates a clear role for computational triage, particularly for SARS-CoV-2, where rapid screening and optimization remain valuable for variant-resilient therapeutic design [2603.04480].

Ab-Affinity addresses a narrower problem than generic antibody–antigen affinity prediction. The training target is a peptide in the spike HR2 region, chosen because it is conserved across SARS-CoV-2 variants and also across broader coronaviruses including SARS and MERS [2603.04480]. That choice makes the system relevant to broad-reactive antibody design, but it also means the model is not a universal antibody–antigen predictor [2603.04480]. A common misconception is to treat it as a dual-encoder system over arbitrary antibody and antigen sequences. In fact, the model takes only the antibody sequence as input, while the target peptide remains fixed by the dataset and task definition [2603.04480].

The paper explicitly positions antibody affinity prediction as more difficult than generic protein–protein interaction prediction because both paratope and epitope regions can be flexible and may include intrinsically disordered regions, for which structural data are sparse [2603.04480]. This motivates a sequence model that learns how amino-acid changes alter affinity, especially in mutational neighborhoods relevant to antibody engineering [2603.04480]. This emphasis differs from benchmarking frameworks that treat affinity as a broader cross-target ranking problem, such as AbRank [2506.17857], and from antigen-specific ranking formulations that adapt to contextual demonstrations at inference time, such as AbICL [2607.05846].

## 2. Model architecture and representation strategy

Ab-Affinity is a BERT-style transformer encoder based on ESM-2; the paper states that the architecture is “based on BERT, as implemented in ESM-2” [2603.04480]. It consists of \(N\) sequential encoder blocks, each containing multi-head self-attention followed by feed-forward layers, and the authors experimented with \(N=6,12,33\), corresponding to models of about 8M, 35M, and 650M parameters, respectively [2603.04480]. The final and best model is the 33-layer version fine-tuned from pretrained ESM-2 weights [2603.04480].

The antibody is represented as an scFv amino-acid sequence, which in this setting effectively combines variable heavy and variable light information into one sequence [2603.04480]. The model processes this amino-acid token sequence through the ESM-2 encoder, and the output of the last encoder layer is taken as the sequence representation, described by the authors as “the embedding” [2603.04480]. Depending on model size, the embedding dimension is 320, 480, or 1280 [2603.04480]. On top of this embedding, the model places a single fully connected layer to regress binding affinity [2603.04480].

The prediction target is the log-transformed dissociation constant, specifically \(\log_{10}(K_d)\) [2603.04480]. Since lower \(K_D\) indicates tighter binding, lower predicted values correspond to stronger affinity [2603.04480]. The training loss is Mean Squared Error, and model selection uses the Pearson correlation coefficient on the validation set [2603.04480]. The paper does not report more elaborate heads, alternative pooling strategies, explicit antigen inputs, or non-MSE objectives [2603.04480].

Architecturally, the novelty is not a new transformer block but the specialization of a large pretrained protein language model to a high-throughput SARS-CoV-2 antibody engineering dataset [2603.04480]. This places Ab-Affinity in a different category from sequence-only dual-stream antigen–antibody regressors such as DuaDeep-SeqAffinity [2512.22007], from structure-conditioned inverse-folding affinity rankers such as SimBinder-IF [2512.17815], and from listwise or in-context ranking frameworks such as AbLWR [2604.11272] and AbICL [2607.05846].

## 3. Dataset, preprocessing, and training protocol

The training dataset comes from Engelhart et al. and contains 104,972 antibody variants generated by introducing one, two, or three amino-acid changes into antibodies derived from three seed candidates identified from a phage display library [2603.04480]. The seed components are described as Ab-14-VH and Ab-14-VL, Ab-91-VH, and Ab-95-VH and Ab-95-VL [2603.04480]. Each variant was assayed against the HR2 peptide, and each interaction had three biological replicate measurements of \(K_D\), estimated by an indirect competitive binding assay [2603.04480].

Preprocessing is unusually specific. For each antibody–antigen pair, the authors took the arithmetic mean of the two closest \(K_D\) values and discarded the third in order to reduce outlier effects [2603.04480]. Antibodies with all three replicate values missing were removed [2603.04480]. After preprocessing, 71,834 unique antibodies remained for training and evaluation, and the labels were then log-transformed to produce \(\log_{10}(K_d)\) as the regression target [2603.04480].

The split strategy was 85% training and 15% validation, with the distribution of affinity values maintained [2603.04480]. The paper also refers repeatedly to a held-out test set used for comparison plots, but the exact proportions or construction of that test set are not clearly described in the methods section [2603.04480]. This underreporting is one of the main methodological limitations of the study. What is clear is that the best checkpoint in each setup was selected by Pearson correlation on the validation set [2603.04480].

Optimization used Adam with batch size 128 for 100 epochs on four NVIDIA A100 80GB GPUs [2603.04480]. The encoder was fine-tuned from pretrained ESM-2 rather than frozen, and the authors also trained a counterpart model from random initialization to assess the contribution of pretrained protein knowledge [2603.04480]. The paper states that the 33-layer fine-tuned model was best, implying that both model scale and protein-language pretraining matter [2603.04480]. Specific settings such as learning rate, weight decay, warmup, dropout, and scheduler are not reported [2603.04480].

The data regime itself is highly consequential for interpretation. The antibodies occupy mutational neighborhoods around only three seed antibodies, with one to three substitutions each [2603.04480]. This makes the benchmark highly relevant for local affinity optimization, but less clearly informative about de novo generalization far from the seed sequences [2603.04480]. That limitation parallels broader concerns in affinity benchmarking about scaffold locality versus true out-of-distribution generalization, a distinction made explicit in AbRank [2506.17857].

## 4. Empirical performance and comparative evaluation

The principal evaluation metrics are Pearson correlation and Spearman rank correlation between predicted and measured affinity; for downstream classification tasks the paper uses ROC curves and AUC [2603.04480]. In the full held-out test-set comparison, the main baselines are DG-Affinity, ESM-2 embeddings plus linear regression, and AbLang embeddings plus linear regression [2603.04480]. The body text explicitly gives DG-Affinity’s Pearson correlation as 0.194 and states that Ab-Affinity has the highest Pearson and Spearman among the four methods [2603.04480]. The exact numeric test-set correlations for ESM-2, AbLang, and Ab-Affinity are not printed in the text, so they are not reproducible from the paper excerpt [2603.04480].

The most concrete quantitative comparison is on the Ab-14-derived subsets 14H and 14L [2603.04480]. The results reported in the paper are as follows.

| Benchmark | Method | Reported result |
|---|---|---|
| 14H | Ab-Affinity | Pearson 0.652, Spearman 0.526 |
| 14H | A2Binder | Pearson 0.642, Spearman 0.553 |
| 14H | ESM-F | Pearson 0.634, Spearman 0.516 |
| 14L | Ab-Affinity | Pearson 0.712, Spearman 0.713 |
| 14L | A2Binder | Pearson 0.683, Spearman 0.688 |
| 14L | ESM-F | Pearson 0.674, Spearman 0.681 |

These results show that Ab-Affinity achieved the best Pearson correlation on both 14H and 14L, and the best Spearman correlation on 14L [2603.04480]. On 14H, its Spearman correlation is slightly below A2Binder, but the paper states that it “closely followed” [2603.04480]. The authors also argue that these Ab-14-specific evaluations understate the broader value of the model because Ab-Affinity performed strongly when evaluated across all three seed antibody families [2603.04480].

The comparative interpretation is narrow but important. DG-Affinity, although previously strong on another antibody dataset, performed poorly on this SARS-CoV-2-specific task [2603.04480]. Generic protein pretraining via ESM-2 and antibody-specific pretraining via AbLang remained reasonably competitive, but supervised specialization to the HR2 affinity dataset yielded the best reported correlation [2603.04480]. This suggests that target-specific supervised fine-tuning can dominate generic pretrained embeddings in local antibody optimization settings [2603.04480].

## 5. Embedding geometry, downstream tasks, and interpretability

A major claim of Ab-Affinity is that fine-tuning reorganizes antibody sequence space into a functionally meaningful latent representation [2603.04480]. Using t-SNE with perplexity 200 on the learned embeddings, the paper reports that Ab-Affinity arranges antibodies along a smooth affinity gradient, whereas raw ESM-2 embeddings do not clearly order antibodies by \(\log K_d\) [2603.04480]. This is presented as evidence that affinity supervision reshapes the latent space rather than merely improving scalar regression [2603.04480].

The same embeddings are used for two downstream classification tasks. One assigns antibodies to High, Medium, and Low affinity classes; the other predicts whether a mutant binds more strongly than its seed antibody [2603.04480]. In both cases, classifiers built on Ab-Affinity embeddings yield ROC and AUC behavior reported as much better than classifiers built on raw ESM-2 embeddings, although exact AUC values are not given in the text excerpt [2603.04480]. This suggests that the representation carries rank- and class-relevant information beyond the final regressed \(\log_{10}(K_d)\) value.

The paper also analyzes transformer attention maps using the contact-analysis method of Rao et al. [2603.04480]. It compares strong binders, defined as \(\log K_d < 0.5\), to weak binders, defined as \(\log K_d > 5.5\), and reports that the largest differences in attention-derived contact maps tend to occur in CDR-H1, CDR-H2, CDR-L1, or adjacent regions [2603.04480]. This is biologically plausible because CDRs mediate antigen binding [2603.04480]. The result is presented as interpretability evidence that the model’s internal focus aligns with known antibody binding sites [2603.04480].

A further analysis examines thermostability. The authors assembled a small external dataset of experimentally determined thermostability for 26 SARS-CoV-2 antibodies from two prior studies and report that Ab-Affinity embeddings separate antibodies into clusters with similar thermostability more clearly than raw ESM-2 embeddings in t-SNE projections [2603.04480]. The paper does not provide a rigorous predictive benchmark for this claim, and the sample size is very small [2603.04480]. A plausible implication is that affinity fine-tuning induces representations that capture some broader biophysical variation, but the evidence remains exploratory rather than conclusive.

## 6. Position within antibody-affinity modeling

Ab-Affinity occupies a specific niche within a rapidly diversifying literature on antibody affinity modeling. It differs from broad ranking benchmarks such as AbRank, which reframes affinity prediction as pairwise ranking across heterogeneous antibodies, antigens, and assay types [2506.17857]. It also differs from antigen-specific contextual rankers such as AbICL, which use support demonstrations to adapt ranking behavior without gradient updates for a given antigen campaign [2607.05846]. Ab-Affinity instead addresses a single conserved SARS-CoV-2 HR2 peptide and predicts a continuous \(\log_{10}(K_d)\) target from antibody sequence alone [2603.04480].

This specialization has both strengths and constraints. Relative to regression-averse frameworks such as AbRank [2506.17857] or listwise ranking systems such as AbLWR [2604.11272], Ab-Affinity retains a direct thermodynamic-like regression target, which is useful when labels are dense, internally consistent, and assay-matched. Its success therefore supports the view that absolute-value regression can still be effective when the antigen is fixed and the experimental design is highly controlled [2603.04480]. By contrast, methods designed for heterogeneous multi-antigen settings often emphasize ranking because cross-assay calibration is difficult [2506.17857].

Ab-Affinity also differs from structure-first approaches. The AlphaFold 3 study on anti-CD47 antibodies evaluates structure generation followed by external MM/GBSA scoring, and concludes that AF3 is useful primarily as an upstream structural modeler rather than a direct affinity predictor [2511.14676]. Ab-Affinity bypasses this route by learning from sequence only, motivated partly by the scarcity of reliable structural data for flexible antibody–antigen regions [2603.04480]. Conversely, a plausible implication is that Ab-Affinity’s scope is strongest where target specialization and dense mutational affinity data are available, rather than in broad complex-level structure generalization.

From a methodological lineage perspective, Ab-Affinity can also be situated against earlier experimental work on sequence–affinity landscapes. Tite-Seq established that massively parallel titration curves can recover physical \(K_D\) values for antibody variants and thereby disentangle affinity from confounding expression and stability effects [1601.02160]. Later analysis of Tite-Seq data showed that binding free-energy landscapes contain substantial epistasis and beneficial interaction effects [1712.04000]. Ab-Affinity inherits the same general engineering logic—learning affinity-relevant structure from mutational sequence data—but implements it through large-scale protein language model fine-tuning rather than direct biophysical landscape inference [2603.04480].

## 7. Limitations, interpretation, and significance

The paper’s limitations are explicit in several places and inferable in others. First, the supervised task concerns antibodies against one selected HR2 peptide rather than arbitrary antigens, so Ab-Affinity is not a universal antibody–antigen affinity predictor [2603.04480]. Claims about broader SARS-CoV-2 or pan-antigen generalization should therefore be interpreted narrowly [2603.04480]. Second, the dataset consists of mutational neighborhoods around only three seed antibodies, which makes the model particularly relevant for local optimization but leaves de novo extrapolation uncertain [2603.04480]. Third, the train/validation/test protocol is incompletely specified, which complicates precise assessment of held-out novelty [2603.04480].

Reporting incompleteness extends to implementation details and some comparison metrics. The main full-test-set numeric correlations are not fully printed, some hyperparameters are missing, and pretrained checkpoint details are only partially described [2603.04480]. These omissions do not negate the reported results, but they limit reproducibility. More broadly, the paper offers strong evidence for target-specialized mutational affinity modeling, not for universal antibody design.

Within that scope, the significance is clear. Ab-Affinity shows that supervised fine-tuning of a large protein language model on a high-throughput SARS-CoV-2 mutational affinity dataset can produce an antibody sequence representation that is predictive of \(\log_{10}(K_d)\), useful for ranking and classification, and partly interpretable through latent geometry and attention behavior [2603.04480]. Its practical value lies in reducing experimental screening burden for SARS-CoV-2 antibody optimization, while its scientific value lies in demonstrating that large protein language models can be specialized into binding-aware sequence encoders for a fixed antigenic target [2603.04480]. A plausible broader implication is that future antibody-affinity systems may increasingly combine large pretrained sequence models with tightly scoped, target-specific supervision rather than relying only on universal architectures or antigen-agnostic training.

Source: https://www.emergentmind.com/topics/ab-affinity