AI Insight
Researchers developed IgScope, an artificial intelligence tool that evaluates how human-like therapeutic antibodies are at multiple levels of resolution, from entire chains down to individual amino acid residues. The 23-million-parameter language model can specifically assess the humanness of complementarity-determining regions (CDRs), which are critical for antibody function but poorly evaluated by existing methods. Testing on 2.38 million antibody sequences showed IgScope accurately detects non-human residues and correlates strongly with clinical immune responses, while revealing that antibodies from four generative AI models appear human overall but have distinctly non-human CDRs.
Why it matters
This tool addresses a critical safety concern in antibody drug development, as non-human-like antibodies can trigger unwanted immune responses that reduce treatment effectiveness or cause adverse reactions. IgScope provides quality control for both traditional antibody humanization and emerging AI-generated antibody designs, potentially accelerating therapeutic development while improving patient safety.
Understand the Science
⚠️ Preprint – Noch nicht peer-reviewed
Dieser Artikel wurde noch nicht von unabhängigen Experten begutachtet. Die Ergebnisse sind vorläufig und sollten mit Vorsicht interpretiert werden.
Antibodies recognized as foreign can elicit anti-drug antibodies that compromise efficacy and safety, making humanness scoring—estimating how human-like a sequence appears—an early, repeated step in antibody development. Widely used scores reduce a variable domain to one framework-dominated number, leaving the complementarity-determining regions (CDRs)—grafted during humanization, diversified by generative models—weakly supervised and rarely audited. Here we introduce IgScope, a 23-million-parameter antibody language model trained from scratch that scores humanness at chain, IMGT-region and residue resolution; a region-isolated attention module confines attention to within-region pairs, and chimeric-sequence training supplies region-level supervision. On 2.38 million held-out chains, masking all framework tokens leaves CDR accuracy unchanged, and back-mutated murine residues in human frameworks are detected at 93.4% accuracy. At chain level, IgScope matches the strongest published correlation with clinical anti-drug-antibody incidence (r = $-$0.567, n = 217) and outperforms a same-protocol recomputation. Region-level scoring exposes what chain-level scoring conceals: antibodies designed de novo by four generative models pass whole-chain humanness (0.966–0.988) while their CDRs score far below natural (P $<$ 1e-30) in every chain class. Quality control turns the same resolution to the mirror-image question: in humanized-mouse platforms, whose domains are human by design, IgScope scores the domain as human and, in HUGO-Nano chimeric VHHs, isolates only the FR2 signal of four camelid hallmark residues deliberately retained to stabilise the VHH. IgScope gives antibody humanization a verified foundation and generative design an audit.
Source: Multi-resolution humanness scoring of antibodies via a region-isolated language model