What Is Miniprotein? Definition and What Research Reports

Plain definition

A miniprotein is a small polypeptide that folds into a defined, stable three-dimensional structure — in other words, a real protein that happens to be very short. Where a classical protein may run to hundreds of amino acids and a conventional peptide is often a short, flexible chain with little fixed shape, a miniprotein sits in between: long enough to form helices, sheets, loops and a hydrophobic core, short enough to be made synthetically or expressed easily in bacteria. In current usage the word most often refers to de novo designed miniproteins: sequences invented by computational protein design rather than borrowed from nature, built specifically to grip a chosen molecular target.

This page is for educational purposes only and is not medical advice; consult a licensed physician about any medical question. Nothing here describes how any compound should be used, and none of the papers summarised below are consumer-facing instructions.

What class of molecule is it?

Miniproteins are polypeptides — chains of amino acids joined by peptide bonds — and so belong to the same broad chemical family as peptides, hormones, antibodies and enzymes. The distinction is structural rather than chemical. Field usage generally applies the label to folded polypeptides in the range of roughly thirty to a hundred residues, often stabilised by a compact hydrophobic core, by disulfide bonds, or by both. Because the fold is pre-organised, the binding surface is held in place before the target is encountered, which is the property protein engineers are usually after.

Miniproteins come from three broad places:

• Natural scaffolds. Small folded domains found in nature — protease inhibitors, defensins, zinc-finger-like folds and similar motifs — have long been used as starting templates for engineering.
• Single-domain antibody fragments. A review of nanobodies described them as robust miniprotein binders used across biomedicine, covering their structure, engineering and applications (PMID 36754285).
• Computational de novo design. Sequences and backbones generated from scratch by design software, then tested in the laboratory. Most recent miniprotein papers fall into this category.

How the term is used in peptide research

In practice, "miniprotein" is a functional label. Researchers use it when the molecule is too small and too synthetic-feeling to be called a protein drug, but too structured to be described as a peptide. Typical uses of the word in the literature include:

1. Miniprotein binder — a designed molecule whose job is to attach tightly and selectively to a target surface.
2. Miniprotein inhibitor — a binder that blocks a target's function, such as receptor engagement or complex assembly.
3. Miniprotein agonist or antagonist — a binder that switches receptor signalling on or off.
4. Conditionally active miniprotein — a binder whose activity is masked until a trigger removes the mask.

Because the folds are small and rigid, miniproteins are frequently discussed alongside stability engineering. One earlier study examined engineering protein stability with atomic precision in a monomeric miniprotein, treating the small fold as a testbed for understanding how individual atoms contribute to stability (PMID 28530710).

What the published literature reports

The bulk of the miniprotein literature is laboratory work: designing candidates computationally, expressing them, measuring binding affinity and selectivity, solving structures, and testing function in cells or animal models. Several representative reports are summarised below.

Designed inhibitors of cell-surface receptors

A 2023 report described the de novo design of miniprotein inhibitors that researchers characterised as highly selective for the integrins αvβ6 and αvβ8 (PMID 37704610). Selectivity is emphasised in that work because integrin family members share closely related binding pockets.

Antiviral miniproteins

A 2025 study reported that designed miniproteins potently inhibited MERS-CoV and protected against it in the models tested (PMID 40450691). Separately, researchers described the computational design of what the study characterised as an ultrapotent deltacoronavirus miniprotein inhibitor (PMID 42054371).

Antibacterial targets

Two related reports described de novo designed, high-affinity miniprotein binders targeting a Francisella tularensis virulence factor, first as a preprint (PMID 40631088) and then in a peer-reviewed journal (PMID 41117072). A further preprint described the de novo design of miniprotein inhibitors of bacterial adhesins, the surface molecules bacteria use to stick to host tissue (PMID 40894640).

Immune and signalling targets

One report described the design of miniprotein inhibitors targeting complement C9 with the stated aim of blocking membrane attack complex assembly (PMID 41813685). Work on G protein-coupled receptors has been reported both as a preprint on miniprotein agonists and antagonists targeting GPCRs (PMID 40501737) and as a journal article on the de novo design of miniproteins targeting GPCRs (PMID 42168559).

Control strategies

A 2025 study described the de novo design of peptide masks that enabled rapid generation of conditionally active miniprotein binders — that is, binders intended to stay inactive until unmasked (PMID 41284265).

Summary table

Miniprotein, peptide and protein: how the words differ

These boundaries are conventions, not strict definitions, and different laboratories draw them in slightly different places.

Adverse Events: What Studies Report

The verified literature summarised on this page consists of computational design papers, biochemical and structural characterisation, cell-based assays and animal work. None of these reports were human clinical trials, and none of them reported human adverse-event rates, human dosing schedules or tolerability outcomes in people. Where protection or inhibition was reported — for example in the MERS-CoV study (PMID 40450691) — the findings belonged to laboratory and preclinical models. Readers evaluating claims about miniproteins should note that a designed binder with strong laboratory affinity has not, on that basis alone, been shown to be safe or effective in humans.

Limitations of the current literature

• Several of the reports cited above are preprints that had not completed peer review at the time of posting (PMID 40894640, PMID 40501737, PMID 40631088).
• Design papers typically report the successful candidates; the failure rate of a design campaign is not always visible from the headline result.
• Binding affinity, selectivity and in vitro potency are laboratory endpoints and do not translate automatically into clinical outcomes.
• The word "miniprotein" is used loosely across sub-fields, so two papers using the term may be describing structurally quite different molecules.

References

• De novo design of highly selective miniprotein inhibitors of integrins αvβ6 and αvβ8 (Nature Communications, 2023)
• De Novo Design of High-Affinity Miniprotein Binders Targeting Francisella Tularensis Virulence Factor (Angewandte Chemie International Edition, 2025)
• Designed miniproteins potently inhibit and protect against MERS-CoV (Cell Reports, 2025)
• De Novo Design of Miniprotein Inhibitors of Bacterial Adhesins (bioRxiv, 2025)
• De novo design of miniproteins targeting GPCRs (Nature, 2026)
• Design of miniprotein inhibitors targeting complement C9 to block membrane attack complex assembly (Nature Communications, 2026)
• De novo design of high-affinity miniprotein binders targeting Francisella tularensis virulence factor (bioRxiv, 2025)
• De novo design of miniprotein agonists and antagonists targeting G protein-coupled receptors (bioRxiv, 2025)
• Computational design of an ultrapotent deltacoronavirus miniprotein inhibitor (PNAS, 2026)
• De Novo Design of Peptide Masks Enables Rapid Generation of Conditionally-Active Miniprotein Binders (Journal of the American Chemical Society, 2025)
• Nanobodies: Robust miniprotein binders in biomedicine (Advanced Drug Delivery Reviews, 2023)
• Engineering protein stability with atomic precision in a monomeric miniprotein (Nature Chemical Biology, 2017)