An interpretable open platform for sequence-based antibody developability prediction

bioRxiv · 11 Nov 2026 · Preprint · Open access

Nguyen, Kothiwal + 6 more

The abstract is read at the publisher; the summary is JClub's.

DELPHI, an open antibody developability prediction platform, achieves mean AUC 0.959 for polyreactivity and 0.933 for SEC using CDR H3-cluster cross-validation.

Why it matters: Sequence-based developability is increasingly predictable, but accessible software and retrainable models are uncommon, limiting reproducible screening and hypothesis generation from antibody sequence alone.
What they did: DELPHI trains and compares 25 language-model and classifier combinations, uses CDR H3-cluster cross-validation to reduce sequence-similarity leakage, evaluates performance versus labelled training-set size, and provides residue-level attributions.
The result: Models trained on in-house polyreactivity and SEC data transfer to a 246,293-antibody public library (AUC 0.950 with the deployed model), enabling pre-assay candidate ranking and residue-level engineering hypotheses with retraining for new assays.

AI summary of the abstract · may contain errors · not medical advice · read the paper.

Open access

Free. We’ll email you when a new paper cites this one or any other paper you follow on JClub — at most one email a day. Every email has a one-tap unsubscribe.

Defining the Antigenic Topology and Prospective Binding Breadth of Vaccination-induced SARS-CoV-2 Neutralizing Antibodies
bioRxiv · 11 Nov 2026 · Preprint
AI summary · may contain errors
Early plasmablast-derived human monoclonal antibodies from primary mRNA vaccination show potent spike RBD/NTD neutralization, but variant-associated mutations more often impact antibodies to immunodominant regions than those targeting a conserved NTD hydrophobic pocket.
- Open access
Open
Share
Post on X Post on Bluesky Share on LinkedIn
https://jclub.com.au/a/10.64898/2026.03.03.709398
Open in JClub
In situ profiling of plasma cell clonality with image-based single-cell transcriptomics.
Nature Methods · 10 Nov 2026 · Journal Article
AI summary · may contain errors
BCR-MERFISH accurately identifies plasma cell V gene usage and clonality in mice, including settings with restricted or native plasma cell diversity, enabling in situ receptor profiling.
- Open access
- 4 cites
Open
Share
Post on X Post on Bluesky Share on LinkedIn
https://jclub.com.au/a/43009623
Open in JClub
Use of Antibody Structural Information in Disease Prediction Models Reveals Antigen Specific B Cell Receptor Sequences in Bulk Repertoire Data
bioRxiv · 10 Nov 2026 · Preprint
AI summary · may contain errors
Structural information from IgH CDR1–3 plus antibody isotype identifies antigen-specific B cell receptor groups in bulk IgH data, improving food allergen prediction and matching HIV prediction performance.
- Open access
Open
Share
Post on X Post on Bluesky Share on LinkedIn
https://jclub.com.au/a/10.1101/2024.12.10.627792
Open in JClub

Journals this month

All journals →

Topics moving, week to 13 Nov 2026

All topics →

By field

All journals by field →

An interpretable open platform for sequence-based antibody developability prediction

Related in JClub

Defining the Antigenic Topology and Prospective Binding Breadth of Vaccination-induced SARS-CoV-2 Neutralizing Antibodies

In situ profiling of plasma cell clonality with image-based single-cell transcriptomics.

Use of Antibody Structural Information in Disease Prediction Models Reveals Antigen Specific B Cell Receptor Sequences in Bulk Repertoire Data

Browse JClub