A deep learning framework (CreoPep) for target-specific design and optimization of conotoxin peptides

Conotoxins are small, disulfide-rich peptides that display exceptional affinity and selectivity for ion channels and receptors, making them valuable templates for therapeutic development. However, their optimization remains challenging due to the limited diversity of naturally occurring variants and the labor-intensive nature of conventional engineering strategies. Here, we present CreoPep, a deep learning-based generative framework specifically developed to design and optimize conotoxins targeting defined receptors. CreoPep integrates masked language modeling with a progressive masking scheme and employs an augmentation pipeline that combines physics-based energy screening with temperature-controlled multinomial sampling. This enables the generation of structurally and functionally diverse peptide variants while retaining essential pharmacological features. Structural analysis shows that CreoPep-generated variants adopt both conserved and previously unobserved binding modes, including disulfide-deficient forms. Together, these findings establish CreoPep as a powerful computational-experimental framework for the rational design of conotoxin-based peptides and provide a foundation for extending similar approaches to other peptide families. Conotoxins are disulfide-rich therapeutic peptides with high affinity and selectivity for ion channels, yet their optimization is hindered by limited sequence diversity and laborious engineering. Here, the authors introduce CreoPep, a deep learning-based generative framework that integrates a progressive masking strategy and an augmentation pipeline that combines physics-based energy screening with temperature-controlled multinomial sampling, rationally designing and generating diverse and potent conotoxin variants.

• Publication year

  2026

• Venue

  Communications Chemistry

• Publication date

  2026-01-09

• Fields of study

  Chemistry, Medicine, Computer Science

• Identifiers
  DOI 10.1038/s42004-025-01885-5PMID 41513959PMCID 12894894
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar, PubMed

• No claims are published for this paper.

• No concepts are published for this paper.

• Advance in peptide-based drug development: delivery platforms, therapeutics and vaccines

  2025influential reference
• A unified deep framework for peptide–major histocompatibility complex–T cell receptor binding prediction

  2025cited by this paper
• De novo designed proteins neutralize lethal snake venom toxins

  2025cited by this paper
• De novo design of self-assembling peptides with antimicrobial activity guided by deep learning.

  2025cited by this paper
• Accurate structure prediction of biomolecular interactions with AlphaFold 3

  2024influential reference
• Design of target specific peptide inhibitors using generative deep learning and molecular dynamics simulations

  2024cited by this paper
• RLpMIEC: High-Affinity Peptide Generation Targeting Major Histocompatibility Complex-I Guided and Interpreted by Interaction Spectrum-Navigated Reinforcement Learning

  2024cited by this paper
• Spike deep mutational scanning helps predict success of SARS-CoV-2 clades

  2024cited by this paper
• Conotoxins Targeting Voltage-Gated Sodium Ion Channels

  2024cited by this paper
• De novo design of protein structure and function with RFdiffusion

  2023cited by this paper
• Computational design of α-conotoxins to target specific nicotinic acetylcholine receptor subtypes.

  2023cited by this paper
• Cyclic Peptides for Drug Development.

  2023cited by this paper
• AmberTools

  2023cited by this paper
• Pan-Peptide Meta Learning for T-cell receptor–antigen binding recognition

  2023cited by this paper
• Designing antimicrobial peptides using deep learning and molecular dynamic simulations

  2023cited by this paper
• Identification of potent antimicrobial peptides via a machine-learning pipeline that mines the entire space of peptide sequences

  2023cited by this paper
• De novo design of high-affinity binders of bioactive helical peptides

  2023cited by this paper
• Robust deep learning based protein sequence design using ProteinMPNN

  2022cited by this paper
• Controllable protein design with language models

  2022cited by this paper
• Therapeutic peptides: current applications and future directions

  2022cited by this paper
• Evolutionary-scale prediction of atomic level protein structure with a language model

  2022cited by this paper
• Derivatization with fatty acids in peptide and protein drug discovery

  2022cited by this paper
• High-resolution de novo structure prediction from primary sequence

  2022influential reference
• Accurate prediction of protein structures and interactions using a 3-track neural network

  2021cited by this paper
• Trends in peptide drug discovery

  2021cited by this paper
• Structure and gating mechanism of the α7 nicotinic acetylcholine receptor.

  2021cited by this paper
• Structure of human Cav2.2 channel blocked by the painkiller ziconotide

  2021cited by this paper
• Highly accurate protein structure prediction with AlphaFold

  2021influential reference
• A deep-learning framework for multi-level peptide–protein interaction prediction

  2021cited by this paper
• ProtTrans: Towards Cracking the Language of Life’s Code Through Self-Supervised Learning

  2021cited by this paper
• ProtTrans: Towards Cracking the Language of Life’s Code Through Self-Supervised Deep Learning and High Performance Computing

  2020cited by this paper
• Denoising Diffusion Probabilistic Models

  2020cited by this paper
• FoldX 5.0: working with RNA, small molecules and a new graphical interface

  2019cited by this paper
• Molecular basis for pore blockade of human Na+ channel Nav1.2 by the μ-conotoxin KIIIA

  2019cited by this paper
• Conotoxins: Chemistry and Biology.

  2019influential reference
• HPEPDOCK: a web server for blind peptide–protein docking based on a hierarchical algorithm

  2018cited by this paper
• Discovery of peptide ligands through docking and virtual screening at nicotinic acetylcholine receptor homology models

  2017cited by this paper
• Role of CysI–CysIII Disulfide Bond on the Structure and Activity of α-Conotoxins at Human Neuronal Nicotinic Acetylcholine Receptors

  2017cited by this paper
• Peptide therapeutics: current status and future directions.

  2015cited by this paper
• Discovery, synthesis, and structure-activity relationships of conotoxins.

  2014cited by this paper
• Conus Venom Peptide Pharmacology

  2012cited by this paper
• ConoServer: updated content, knowledge, and discovery tools in the conopeptide database

  2011influential reference
• Blockade of Neuronal α7-nAChR by α-Conotoxin ImI Explained by Computational Scanning and Energy Calculations

  2011cited by this paper
• Rosetta FlexPepDock web server—high resolution modeling of peptide–protein interactions

  2011cited by this paper
• How significant is a protein structure similarity with TM-score = 0.5?

  2010cited by this paper
• Inhibition of Neuronal Nicotinic Acetylcholine Receptor Subtypes by α-Conotoxin GID and Analogues*

  2009cited by this paper
• Ziconotide: neuronal calcium channel blocker for treating severe chronic pain.

  2004cited by this paper
• Conus venoms: a rich source of novel ion channel-targeted peptides.

  2004cited by this paper
• Isolation, Structure, and Activity of GID, a Novel α4/7-Conotoxin with an Extended N-terminal Sequence*

  2003cited by this paper
• Neuronal calcium channel antagonists. Discrimination between calcium channel subtypes using omega-conotoxin from Conus magus venom.

  1987influential reference
• A simple method for displaying the hydropathic character of a protein.

  1982cited by this paper

• No citing papers are available for this paper.