Antibodies are crucial proteins produced by the immune system to eliminate harmful foreign substances and have become pivotal therapeutic agents for treating human diseases.

To accelerate therapeutic antibody discovery, computational methods, especially machine learning, have attracted considerable interest for predicting specific interactions between antibody candidates and target antigens such as viruses and bacteria.

Antibodies are crucial proteins produced by the immune system to eliminate harmful foreign substances and have become pivotal therapeutic agents for treating human diseases. To accelerate the discovery of antibody therapeutics, there is growing interest in constructing language models using antibody sequences. However, the applicability of pre-trained language models for antibody discovery has not been thoroughly evaluated due to the scarcity of labeled datasets. To overcome these limitations, we introduce AVIDa-SARS-CoV-2, a dataset featuring the antigen-variable domain of heavy chain of heavy chain antibody (VHH) interactions obtained from two alpacas immunized with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) spike proteins. AVIDa-SARS-CoV-2 includes binary labels indicating the binding or non-binding of diverse VHH sequences to 12 SARS-CoV-2 mutants, such as the Delta and Omicron variants. Furthermore, we release VHHCorpus-2M, a pre-training dataset for antibody language models, containing over two million VHH sequences. We report benchmark results for predicting SARS-CoV-2-VHH binding using VHHBERT pre-trained on VHHCorpus-2M and existing general protein and antibody-specific pre-trained language models. These results confirm that AVIDa-SARS-CoV-2 provides valuable benchmarks for evaluating the representation capabilities of antibody language models for binding prediction, thereby facilitating the development of AI-driven antibody discovery.

To increase the accuracy of the analysis, deep sequencing data were filtered; target sequences with deep sequencing read counts below 200 and background indel frequencies above 8% were excluded as similarly performed previously (21). DNase-sequencing (DNase-seq) narrow peak data from ENCODE (36) were used to calculate chromatin accessibility as previously described (21). For each target site, 23 bases of the PAM plus protospacer sequence were aligned to the hg19 human reference genome using bowtie (41). Only the target sites that overlapped with DNase-seq narrow peaks were considered as DNase I hypersensitive target sites. We divided the Endo_Cas9 dataset into paired subsets by stratified random sampling from strata of DHS and non-DHS sites so that a similar ratio of DHS/non-DHS sites was assigned to each subset.