Back

Accelerating multi-objective VHH discovery via integrated high-throughput selection and AlphaFold3-guided structure prediction

Overath, M. D.; Thumtecho, S.; Rivera-de-Torre, E.; Benard Valle, M.; Wolff, D. S.; Grahadi, R.; Björnsson, K. H.; Rygaard, A. S. H.; Hofmann, N.; Ledergerber, J.; Ljungars, A.; Laustsen, A. H.; Olsson, S.; Fryer, T. J.; Jenkins, T. P.

2026-01-20 bioengineering
10.64898/2026.01.19.700436 bioRxiv
Show abstract

Discovering therapeutic antibodies that bind multiple related targets with high affinity and favourable biophysical properties remains challenging and resource intensive. For snakebite antivenoms, this challenge is critical as treatments must neutralise toxins across multiple snake species. We developed a pipeline combining high-throughput yeast screening, deep sequencing, and AlphaFold3 structure prediction to rapidly identify poly-specific variable domains of heavy-chain-only antibodies (VHHs) against long-chain -neurotoxins. Multiplexed yeast display screening generated a dataset of diverse candidates with varying binding specificities. AlphaFold3-generated VHH-toxin complex predictions enabled structure-based prioritisation that accurately predicted poly-specific binders targeting conserved epitopes across multiple toxins. These structural insights enabled computational optimisation of both affinity and solubility without disrupting target recognition. Experimental validation confirmed improved variants maintained broad specificity across toxins. This integrated approach accelerates multi-objective antibody discovery by predicting which candidates will bind multiple targets before extensive laboratory testing, providing a generalisable strategy applicable beyond antivenoms to any therapeutic requiring broad target coverage.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

1
Nature Communications
○ 5641 papers in training set
Top 11%
17.9%
Similar papers in this journal
2
Advanced Science
○ 286 papers in training set
Top 0.4%
9.4%
Similar papers in this journal
3
Science Advances
● 1243 papers in training set
Top 3%
7.0%
Similar papers in this journal
4
Proceedings of the National Academy of Sciences
● 2444 papers in training set
Top 11%
4.7%
Similar papers in this journal
5
Cell Reports
○ 1498 papers in training set
Top 10%
3.9%
6
eLife
● 5828 papers in training set
Top 32%
3.9%
7
Protein Engineering, Design and Selection
◐ 15 papers in training set
Top 0.1%
3.9%
Similar papers in this journal
50% of probability mass above
8
Communications Biology
○ 993 papers in training set
Top 4%
3.9%
9
Protein Science
○ 246 papers in training set
Top 1%
3.1%
10
Science
● 477 papers in training set
Top 3%
3.1%
11
Journal of the American Chemical Society
● 217 papers in training set
Top 1%
2.3%
12
Structure
○ 193 papers in training set
Top 1%
2.1%
13
PLOS Pathogens
● 820 papers in training set
Top 6%
2.1%
14
Molecular Therapy
○ 81 papers in training set
Top 0.8%
2.1%
15
Nature Biotechnology
○ 172 papers in training set
Top 2%
2.1%
16
Angewandte Chemie International Edition
○ 93 papers in training set
Top 0.8%
1.8%
17
Nature Methods
○ 385 papers in training set
Top 4%
1.7%
18
Cell Systems
○ 201 papers in training set
Top 3%
1.4%
19
Cell Chemical Biology
○ 94 papers in training set
Top 0.9%
1.4%
20
PLOS Computational Biology
● 1863 papers in training set
Top 16%
1.4%
21
mAbs
○ 32 papers in training set
Top 0.4%
1.1%
22
ACS Chemical Biology
● 167 papers in training set
Top 2%
1.1%
23
Neuron
○ 337 papers in training set
Top 4%
1.1%
24
Cell Reports Methods
○ 165 papers in training set
Top 3%
1.1%
25
Journal of Molecular Biology
○ 232 papers in training set
Top 3%
1.0%
26
Nature Biomedical Engineering
○ 47 papers in training set
Top 2%
1.0%
27
ACS Synthetic Biology
● 287 papers in training set
Top 2%
0.8%
28
Nucleic Acids Research
◐ 1281 papers in training set
Top 14%
0.8%
29
Cell Host & Microbe
126 papers in training set
Top 2%
0.8%
30
EMBO Reports
○ 263 papers in training set
Top 7%
0.8%

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.