Harnessing protein folding neural networks for peptide–protein docking

Tomer Tsaban; Julia K. Varga; Orly Avraham; Ziv Ben-Aharon; Alisa Khramushin; Ora Schueler-Furman

doi:10.1038/s41467-021-27838-9

Harnessing protein folding neural networks for peptide–protein docking

Tomer Tsaban, Julia K. Varga, Orly Avraham, Ziv Ben-Aharon, Alisa Khramushin, Ora Schueler-Furman^*

^*Corresponding author for this work

Department of Microbiology and Molecular Genetics

Research output: Contribution to journal › Article › peer-review

75 Scopus citations

Abstract

Highly accurate protein structure predictions by deep neural networks such as AlphaFold2 and RoseTTAFold have tremendous impact on structural biology and beyond. Here, we show that, although these deep learning approaches have originally been developed for the in silico folding of protein monomers, AlphaFold2 also enables quick and accurate modeling of peptide–protein interactions. Our simple implementation of AlphaFold2 generates peptide–protein complex models without requiring multiple sequence alignment information for the peptide partner, and can handle binding-induced conformational changes of the receptor. We explore what AlphaFold2 has memorized and learned, and describe specific examples that highlight differences compared to state-of-the-art peptide docking protocol PIPER-FlexPepDock. These results show that AlphaFold2 holds great promise for providing structural insight into a wide range of peptide–protein complexes, serving as a starting point for the detailed characterization and manipulation of these interactions.

Original language	American English
Article number	176
Journal	Nature Communications
Volume	13
Issue number	1
DOIs	https://doi.org/10.1038/s41467-021-27838-9
State	Published - Dec 2022

Bibliographical note

Funding Information:
We are grateful to Sergey Ovchinnikov, Martin Steinegger and Milot Mirdita, and anyone else that has helped provide notebooks to run AF2. This work was supported, in whole or in part, by the Israel Science Foundation, founded by the Israel Academy of Science and Humanities (grant numbers 717/2017 and 301/2021 to O.S.-F.) and the US-Israel Binational Science Foundation (grant number 2015207). J.K.V. is supported by a Marie Sklodowska-Curie European Training Network Grant #860517.

Publisher Copyright:
© 2022, The Author(s).

Keywords

Amino Acid Sequence
Binding Sites
Models, Molecular
Molecular Docking Simulation
Neural Networks, Computer
Peptides/chemistry
Protein Binding
Protein Conformation, alpha-Helical
Protein Conformation, beta-Strand
Protein Folding
Protein Interaction Domains and Motifs
Protein Interaction Mapping
Proteins/chemistry
Software

Access to Document

10.1038/s41467-021-27838-9

Cite this

@article{af1464fa414a489f8298f78516ccf8b3,

title = "Harnessing protein folding neural networks for peptide–protein docking",

abstract = "Highly accurate protein structure predictions by deep neural networks such as AlphaFold2 and RoseTTAFold have tremendous impact on structural biology and beyond. Here, we show that, although these deep learning approaches have originally been developed for the in silico folding of protein monomers, AlphaFold2 also enables quick and accurate modeling of peptide–protein interactions. Our simple implementation of AlphaFold2 generates peptide–protein complex models without requiring multiple sequence alignment information for the peptide partner, and can handle binding-induced conformational changes of the receptor. We explore what AlphaFold2 has memorized and learned, and describe specific examples that highlight differences compared to state-of-the-art peptide docking protocol PIPER-FlexPepDock. These results show that AlphaFold2 holds great promise for providing structural insight into a wide range of peptide–protein complexes, serving as a starting point for the detailed characterization and manipulation of these interactions.",

keywords = "Amino Acid Sequence, Binding Sites, Models, Molecular, Molecular Docking Simulation, Neural Networks, Computer, Peptides/chemistry, Protein Binding, Protein Conformation, alpha-Helical, Protein Conformation, beta-Strand, Protein Folding, Protein Interaction Domains and Motifs, Protein Interaction Mapping, Proteins/chemistry, Software",

author = "Tomer Tsaban and Varga, {Julia K.} and Orly Avraham and Ziv Ben-Aharon and Alisa Khramushin and Ora Schueler-Furman",

note = "Funding Information: We are grateful to Sergey Ovchinnikov, Martin Steinegger and Milot Mirdita, and anyone else that has helped provide notebooks to run AF2. This work was supported, in whole or in part, by the Israel Science Foundation, founded by the Israel Academy of Science and Humanities (grant numbers 717/2017 and 301/2021 to O.S.-F.) and the US-Israel Binational Science Foundation (grant number 2015207). J.K.V. is supported by a Marie Sklodowska-Curie European Training Network Grant #860517. Publisher Copyright: {\textcopyright} 2022, The Author(s).",

year = "2022",

month = dec,

doi = "10.1038/s41467-021-27838-9",

language = "American English",

volume = "13",

journal = "Nature Communications",

issn = "2041-1723",

publisher = "Nature Publishing Group",

number = "1",

}

TY - JOUR

T1 - Harnessing protein folding neural networks for peptide–protein docking

AU - Tsaban, Tomer

AU - Varga, Julia K.

AU - Avraham, Orly

AU - Ben-Aharon, Ziv

AU - Khramushin, Alisa

AU - Schueler-Furman, Ora

N1 - Funding Information: We are grateful to Sergey Ovchinnikov, Martin Steinegger and Milot Mirdita, and anyone else that has helped provide notebooks to run AF2. This work was supported, in whole or in part, by the Israel Science Foundation, founded by the Israel Academy of Science and Humanities (grant numbers 717/2017 and 301/2021 to O.S.-F.) and the US-Israel Binational Science Foundation (grant number 2015207). J.K.V. is supported by a Marie Sklodowska-Curie European Training Network Grant #860517. Publisher Copyright: © 2022, The Author(s).

PY - 2022/12

Y1 - 2022/12

N2 - Highly accurate protein structure predictions by deep neural networks such as AlphaFold2 and RoseTTAFold have tremendous impact on structural biology and beyond. Here, we show that, although these deep learning approaches have originally been developed for the in silico folding of protein monomers, AlphaFold2 also enables quick and accurate modeling of peptide–protein interactions. Our simple implementation of AlphaFold2 generates peptide–protein complex models without requiring multiple sequence alignment information for the peptide partner, and can handle binding-induced conformational changes of the receptor. We explore what AlphaFold2 has memorized and learned, and describe specific examples that highlight differences compared to state-of-the-art peptide docking protocol PIPER-FlexPepDock. These results show that AlphaFold2 holds great promise for providing structural insight into a wide range of peptide–protein complexes, serving as a starting point for the detailed characterization and manipulation of these interactions.

AB - Highly accurate protein structure predictions by deep neural networks such as AlphaFold2 and RoseTTAFold have tremendous impact on structural biology and beyond. Here, we show that, although these deep learning approaches have originally been developed for the in silico folding of protein monomers, AlphaFold2 also enables quick and accurate modeling of peptide–protein interactions. Our simple implementation of AlphaFold2 generates peptide–protein complex models without requiring multiple sequence alignment information for the peptide partner, and can handle binding-induced conformational changes of the receptor. We explore what AlphaFold2 has memorized and learned, and describe specific examples that highlight differences compared to state-of-the-art peptide docking protocol PIPER-FlexPepDock. These results show that AlphaFold2 holds great promise for providing structural insight into a wide range of peptide–protein complexes, serving as a starting point for the detailed characterization and manipulation of these interactions.

KW - Amino Acid Sequence

KW - Binding Sites

KW - Models, Molecular

KW - Molecular Docking Simulation

KW - Neural Networks, Computer

KW - Peptides/chemistry

KW - Protein Binding

KW - Protein Conformation, alpha-Helical

KW - Protein Conformation, beta-Strand

KW - Protein Folding

KW - Protein Interaction Domains and Motifs

KW - Protein Interaction Mapping

KW - Proteins/chemistry

KW - Software

UR - http://www.scopus.com/inward/record.url?scp=85122690750&partnerID=8YFLogxK

U2 - 10.1038/s41467-021-27838-9

DO - 10.1038/s41467-021-27838-9

M3 - Article

C2 - 35013344

AN - SCOPUS:85122690750

SN - 2041-1723

VL - 13

JO - Nature Communications

JF - Nature Communications

IS - 1

M1 - 176

ER -

Harnessing protein folding neural networks for peptide–protein docking

Abstract

Bibliographical note

Keywords

Access to Document

Other files and links

Fingerprint

Cite this