Design of Peptide-Guided Protein Degraders with Structure-Agnostic Language Models

Brixi, Garyk; Ye, Tianzheng; Palepu, Kalyan; Hong, Lauren; Yudistyra, Vivian; Vincoff, Sophia; Christopher, Jayani; Li, Xinning; Bhat, Suhaas; Monticello, Connor; Lopez‐Barbosa, Natalia; Petersen, Lillian; Zang, Tian-Lai; Liu, Tong; Zhao, Lin; Zhang, Sue; Ponnapati, Manvitha; Tysinger, Emma; Stan, Teodora; Koseki, Sabrina; DeLisa, Matthew P.; Chatterjee, Pranam

doi:10.21203/rs.3.rs-1694183/v1

Cited by 2 publications

(1 citation statement)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…24 In multiple studies, we have demonstrated the suitability of the uAb architecture to incorporate linear guide peptides designed via both structure-and sequence-based algorithms. [24][25][26] We thus employ uAbs to evaluate the binding and degradation capacities of PepPrCLIP-designed peptides on four disease-related substrates, including a transcription factor (ETV6), a membrane protein (AMHR2), and two fusion oncoproteins (EWS-FLI1, PAX3-FOXO1).…”

Section: Pepprclip-generated Uabs Induce Robust Degradation Of Pathog...mentioning

confidence: 99%

De NovoDesign of Peptide Binders to Conformationally Diverse Targets with Contrastive Language Modeling

Bhat

Palepu

Yudistyra

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

Designing binders to target undruggable proteins presents a formidable challenge in drug discovery, requiring innovative approaches to overcome the lack of putative binding sites. Recently, generative models have been trained to design binding proteins from the three-dimensional structure of a target protein alone, but thus exclude design to disordered or conformationally unstable targets. In this work, we provide a generalizable algorithmic framework to design short, target-binding peptide motifs, requiring only the amino acid sequence of the target protein. To do this, we propose a process to generate naturalistic peptide candidates through Gaussian perturbation of the peptidic latent space of the state-of-the-art ESM-2 protein language model, and subsequently screen thesede novolinear sequences for target-selective interaction activity via a CLIP-based contrastive learning architecture. By integrating these generative and discriminative steps, we create aPeptidePrioritization viaCLIP(PepPrCLIP) pipeline and validate highly-ranked, target-specific peptide motifs experimentally via fusion to E3 ubiquitin ligase domains, demonstrating functionally potent degradation of conventionally undruggable targetsin vitro. Overall, our design strategy provides a modular toolkit for designing short binding motifs to any target protein without the reliance on stable and ordered tertiary structure, enabling generation of programmable modulators to undruggable and disordered proteins such as transcription factors and fusion oncoproteins.

show abstract

Section: Pepprclip-generated Uabs Induce Robust Degradation Of Pathog...mentioning

confidence: 99%

De NovoDesign of Peptide Binders to Conformationally Diverse Targets with Contrastive Language Modeling

Bhat

Palepu

Yudistyra

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

AF2BIND: Predicting ligand-binding sites using the pair representation of AlphaFold2

Gazizov,

Lian,

Goverde

et al. 2023

Preprint

View full text Add to dashboard Cite

Predicting ligand-binding sites, particularly in the absence of previously resolved homologous structures, presents a significant challenge in structural biology. Here, we leverage the internal pairwise representation of AlphaFold2 (AF2) to train a model, AF2BIND, to accurately predict small-molecule-binding residues given only a target protein. AF2BIND uses 20 "bait" amino acids to optimally extract the binding signal in the absence of a small-molecule ligand. We find that the AF2 pair representation outperforms other neural-network representations for binding-site prediction. Moreover, unique combinations of the 20 bait amino acids are correlated with chemical properties of the ligand.

show abstract

Design of Peptide-Guided Protein Degraders with Structure-Agnostic Language Models

Cited by 2 publications

References 26 publications

De NovoDesign of Peptide Binders to Conformationally Diverse Targets with Contrastive Language Modeling

De NovoDesign of Peptide Binders to Conformationally Diverse Targets with Contrastive Language Modeling

AF2BIND: Predicting ligand-binding sites using the pair representation of AlphaFold2

Contact Info

Product

Resources

About