Comparative Analysis of Deep Learning‐Based Algorithms for Peptide Structure Prediction

While of primary importance in both the biomedical and therapeutic fields, peptides suffer from a relative lack of dedicated tools to predict efficiently and accurately their 3D structures despite being a crucial step in understanding their physio‐pathological function or designing new drugs. In rec...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:Proteins, structure, function, and bioinformatics
Hlavní autoři: Sauvestre, Clément, Zagury, Jean‐François, Langenfeld, Florent
Médium: Journal Article
Jazyk:angličtina
Vydáno: United States 05.10.2025
Témata:
ISSN:0887-3585, 1097-0134, 1097-0134
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Abstract While of primary importance in both the biomedical and therapeutic fields, peptides suffer from a relative lack of dedicated tools to predict efficiently and accurately their 3D structures despite being a crucial step in understanding their physio‐pathological function or designing new drugs. In recent years, deep‐learning methods have enabled a major breakthrough for the protein 3D structure prediction approaches, allowing to predict protein 3D structures with a near‐experimental accuracy for nearly any protein sequence. This present study aims at confronting some of these new methods (AlphaFold2, RoseTTAFold2, and ESMFold) for the peptides' 3D structure prediction problem and evaluating their performance. All methods produced high‐quality results, but their overall performance is lower as compared to the prediction of protein 3D structures. We also identified a few structural features that impede the ability to produce high‐quality peptide structure predictions. These findings point out the discrepancy that still exists between the protein and peptide 3D structure prediction methods and underline a few cases where the generated peptide structures should be used very cautiously.
AbstractList While of primary importance in both the biomedical and therapeutic fields, peptides suffer from a relative lack of dedicated tools to predict efficiently and accurately their 3D structures despite being a crucial step in understanding their physio-pathological function or designing new drugs. In recent years, deep-learning methods have enabled a major breakthrough for the protein 3D structure prediction approaches, allowing to predict protein 3D structures with a near-experimental accuracy for nearly any protein sequence. This present study aims at confronting some of these new methods (AlphaFold2, RoseTTAFold2, and ESMFold) for the peptides' 3D structure prediction problem and evaluating their performance. All methods produced high-quality results, but their overall performance is lower as compared to the prediction of protein 3D structures. We also identified a few structural features that impede the ability to produce high-quality peptide structure predictions. These findings point out the discrepancy that still exists between the protein and peptide 3D structure prediction methods and underline a few cases where the generated peptide structures should be used very cautiously.
While of primary importance in both the biomedical and therapeutic fields, peptides suffer from a relative lack of dedicated tools to predict efficiently and accurately their 3D structures despite being a crucial step in understanding their physio-pathological function or designing new drugs. In recent years, deep-learning methods have enabled a major breakthrough for the protein 3D structure prediction approaches, allowing to predict protein 3D structures with a near-experimental accuracy for nearly any protein sequence. This present study aims at confronting some of these new methods (AlphaFold2, RoseTTAFold2, and ESMFold) for the peptides' 3D structure prediction problem and evaluating their performance. All methods produced high-quality results, but their overall performance is lower as compared to the prediction of protein 3D structures. We also identified a few structural features that impede the ability to produce high-quality peptide structure predictions. These findings point out the discrepancy that still exists between the protein and peptide 3D structure prediction methods and underline a few cases where the generated peptide structures should be used very cautiously.While of primary importance in both the biomedical and therapeutic fields, peptides suffer from a relative lack of dedicated tools to predict efficiently and accurately their 3D structures despite being a crucial step in understanding their physio-pathological function or designing new drugs. In recent years, deep-learning methods have enabled a major breakthrough for the protein 3D structure prediction approaches, allowing to predict protein 3D structures with a near-experimental accuracy for nearly any protein sequence. This present study aims at confronting some of these new methods (AlphaFold2, RoseTTAFold2, and ESMFold) for the peptides' 3D structure prediction problem and evaluating their performance. All methods produced high-quality results, but their overall performance is lower as compared to the prediction of protein 3D structures. We also identified a few structural features that impede the ability to produce high-quality peptide structure predictions. These findings point out the discrepancy that still exists between the protein and peptide 3D structure prediction methods and underline a few cases where the generated peptide structures should be used very cautiously.
Author Zagury, Jean‐François
Langenfeld, Florent
Sauvestre, Clément
Author_xml – sequence: 1
  givenname: Clément
  orcidid: 0000-0002-3814-0514
  surname: Sauvestre
  fullname: Sauvestre, Clément
  organization: Laboratoire GBCM, EA7528, Conservatoire National des Arts et métiers (CNAM) HESAM Université Paris France, Peptinov Paris France
– sequence: 2
  givenname: Jean‐François
  surname: Zagury
  fullname: Zagury, Jean‐François
  organization: Laboratoire GBCM, EA7528, Conservatoire National des Arts et métiers (CNAM) HESAM Université Paris France
– sequence: 3
  givenname: Florent
  orcidid: 0000-0002-3184-3401
  surname: Langenfeld
  fullname: Langenfeld, Florent
  organization: Laboratoire GBCM, EA7528, Conservatoire National des Arts et métiers (CNAM) HESAM Université Paris France, Peptinov Paris France
BackLink https://www.ncbi.nlm.nih.gov/pubmed/41047732$$D View this record in MEDLINE/PubMed
BookMark eNo90M1KxDAUhuEgis6MbrwAyVKE6mmSNu1yHH9hwAF14aqk6alG2qYmqTA7L8Fr9Eqc0dHV4cDLt3jGZLuzHRJyGMNpDMDOemfDqQQQ-RYZxZDLCGIutskIskxGPMmSPTL2_hUA0pynu2RPxCCk5GxEnma27ZVTwbwjnXaqWXrjqa3pBWJP56hcZ7rnr4_Pc-WxotPm2ToTXlpPa-voAvtgKqT3wQ06DA7pwmFldDC22yc7tWo8HmzuhDxeXT7MbqL53fXtbDqPNMtEiGqOdZ1mJUCV1ohMV1IkyKtS54nImUgVS0FBrnVZJkzFLOcVZnL1Mp0Ci_mEHP_urhjeBvShaI3X2DSqQzv4grNEJiwVuVylR5t0KFusit6ZVrll8cexCk5-A-2s9w7r_ySGYm1drK2LH2v-DeYYc0A
Cites_doi 10.1093/nar/gky1047
10.1093/nar/gkad376
10.1021/bi952419l
10.1038/s41586‐024‐07487‐w
10.1093/nar/gkr703
10.1021/jacs.2c12462
10.1016/0263‐7855(90)80070‐V
10.1093/nar/gkm216
10.1093/bib/bbab308
10.1111/j.1742-4658.2009.07360.x
10.1002/prot.26593
10.1093/nar/gkg571
10.1111/j.1399-6576.1997.tb04627.x
10.1016/j.str.2022.11.012
10.3389/fmolb.2021.697586
10.1016/j.toxicon.2018.05.006
10.1101/2023.05.24.542179
10.1093/bioinformatics/btt473
10.1039/D2SC00924B
10.1107/S0021889892009944
10.1038/nature08142
10.1073/pnas.1111471108
10.1038/s41392‐022‐00904‐4
10.1093/nar/28.1.235
10.1002/anie.201501714
10.1002/pep2.24073
10.1016/S0006‐3495(03)74814‐0
10.1107/S0907444913007051
10.7554/eLife.75751
10.1093/nar/gks1169
10.1002/bip.20734
10.1038/s41573‐020‐00135‐8
10.1093/bib/bbae215
10.1038/nsb1203-980
10.1002/prot.26257
10.1038/nsb1101-923
10.1038/s41586‐023‐06832‐9
10.1038/s41573‐021‐00337‐8
10.1186/s13062-015-0103-4
10.1002/prot.26237
10.1038/s41586-021-03828-1
10.1038/nbt.3988
10.1038/s41586-021-03819-2
10.1016/j.sbi.2024.102973
10.1101/622803
10.3390/toxins15100600
10.1021/bi035412
10.1002/(SICI)1097‐0282(199705)41:6<623::AID‐BIP3>3.0.CO;2‐S
ContentType Journal Article
Copyright 2025 The Author(s). PROTEINS: Structure, Function, and Bioinformatics published by Wiley Periodicals LLC.
Copyright_xml – notice: 2025 The Author(s). PROTEINS: Structure, Function, and Bioinformatics published by Wiley Periodicals LLC.
DBID AAYXX
CITATION
NPM
7X8
DOI 10.1002/prot.70049
DatabaseName CrossRef
PubMed
MEDLINE - Academic
DatabaseTitle CrossRef
PubMed
MEDLINE - Academic
DatabaseTitleList PubMed
CrossRef
MEDLINE - Academic
Database_xml – sequence: 1
  dbid: NPM
  name: PubMed
  url: http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?db=PubMed
  sourceTypes: Index Database
– sequence: 2
  dbid: 7X8
  name: MEDLINE - Academic
  url: https://search.proquest.com/medline
  sourceTypes: Aggregation Database
DeliveryMethod fulltext_linktorsrc
Discipline Anatomy & Physiology
Chemistry
Biology
EISSN 1097-0134
ExternalDocumentID 41047732
10_1002_prot_70049
Genre Journal Article
GrantInformation_xml – fundername: Association Nationale de la Recherche et de la Technologie
GroupedDBID -~X
.3N
.GA
05W
0R~
10A
1L6
1OB
1OC
1ZS
33P
3SF
3WU
4.4
4ZD
50Y
50Z
51W
51X
52M
52N
52O
52P
52S
52T
52U
52W
52X
5RE
5VS
66C
702
7PT
8-0
8-1
8-3
8-4
8-5
8UM
930
A03
AAESR
AAEVG
AAHQN
AAMNL
AANLZ
AAONW
AAXRX
AAYCA
AAYXX
AAZKR
ABCQN
ABCUV
ABIJN
ABLJU
ACAHQ
ACCZN
ACFBH
ACGFS
ACIWK
ACPOU
ACPRK
ACXBN
ACXQS
ADBBV
ADEOM
ADIZJ
ADKYN
ADMGS
ADOZA
ADZMN
AEIGN
AEIMD
AEUYR
AEYWJ
AFBPY
AFFPM
AFGKR
AFRAH
AFWVQ
AFZJQ
AGHNM
AGYGG
AHBTC
AHMBA
AITYG
AIURR
AJXKR
ALAGY
ALMA_UNASSIGNED_HOLDINGS
ALVPJ
AMBMR
AMYDB
ATUGU
AUFTA
AZBYB
AZVAB
BAFTC
BDRZF
BFHJK
BHBCM
BLYAC
BMNLL
BNHUX
BROTX
BRXPI
BY8
CITATION
CS3
D-E
D-F
D0L
DCZOG
DPXWK
DR1
DR2
DRFUL
DRSTM
EBS
F00
F01
F04
F5P
G-S
G.N
GNP
GODZA
H.T
H.X
HBH
HGLYW
HHY
HHZ
HZ~
IX1
JPC
KQQ
LATKE
LAW
LC2
LC3
LEEKS
LH4
LITHE
LOXES
LP6
LP7
LUTES
LYRES
MEWTI
MK4
MRFUL
MRSTM
MSFUL
MSSTM
MXFUL
MXSTM
N04
N05
N9A
NF~
NNB
O66
O8X
O9-
P2P
P2W
P2X
P4D
PQQKQ
Q.N
Q11
QB0
QRW
R.K
RNS
ROL
RX1
SUPJJ
UB1
V2E
W8V
W99
WBFHL
WBKPD
WIB
WIH
WIK
WJL
WOHZO
WQJ
WXSBR
WYISQ
XG1
XPP
XV2
ZZTAW
~IA
~WT
ALUQN
NPM
7X8
ID FETCH-LOGICAL-c284t-f3eff68b00d6fee2cd745e3dbc9549246a260a09ccbb52a1293de87ccb2c60213
ISICitedReferencesCount 0
ISICitedReferencesURI http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=001586797900001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN 0887-3585
1097-0134
IngestDate Tue Oct 07 08:12:52 EDT 2025
Tue Oct 07 09:50:15 EDT 2025
Sat Nov 29 07:22:34 EST 2025
IsDoiOpenAccess false
IsOpenAccess true
IsPeerReviewed true
IsScholarly true
Keywords prediction
deep learning
peptide
three‐dimensional structure
alphafold
Language English
License 2025 The Author(s). PROTEINS: Structure, Function, and Bioinformatics published by Wiley Periodicals LLC.
LinkModel OpenURL
MergedId FETCHMERGED-LOGICAL-c284t-f3eff68b00d6fee2cd745e3dbc9549246a260a09ccbb52a1293de87ccb2c60213
Notes ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 23
ORCID 0000-0002-3184-3401
0000-0002-3814-0514
OpenAccessLink https://onlinelibrary.wiley.com/doi/pdfdirect/10.1002/prot.70049
PMID 41047732
PQID 3257526497
PQPubID 23479
ParticipantIDs proquest_miscellaneous_3257526497
pubmed_primary_41047732
crossref_primary_10_1002_prot_70049
PublicationCentury 2000
PublicationDate 2025-10-05
PublicationDateYYYYMMDD 2025-10-05
PublicationDate_xml – month: 10
  year: 2025
  text: 2025-10-05
  day: 05
PublicationDecade 2020
PublicationPlace United States
PublicationPlace_xml – name: United States
PublicationTitle Proteins, structure, function, and bioinformatics
PublicationTitleAlternate Proteins
PublicationYear 2025
References e_1_2_9_31_1
e_1_2_9_52_1
e_1_2_9_50_1
e_1_2_9_10_1
e_1_2_9_35_1
e_1_2_9_12_1
e_1_2_9_33_1
e_1_2_9_14_1
e_1_2_9_39_1
e_1_2_9_16_1
e_1_2_9_37_1
e_1_2_9_18_1
e_1_2_9_41_1
e_1_2_9_20_1
e_1_2_9_22_1
e_1_2_9_45_1
e_1_2_9_24_1
e_1_2_9_43_1
e_1_2_9_8_1
e_1_2_9_6_1
e_1_2_9_4_1
e_1_2_9_2_1
e_1_2_9_26_1
e_1_2_9_49_1
e_1_2_9_28_1
e_1_2_9_47_1
e_1_2_9_30_1
e_1_2_9_51_1
e_1_2_9_11_1
e_1_2_9_34_1
e_1_2_9_13_1
e_1_2_9_32_1
e_1_2_9_15_1
e_1_2_9_38_1
e_1_2_9_17_1
e_1_2_9_36_1
e_1_2_9_19_1
e_1_2_9_42_1
e_1_2_9_40_1
e_1_2_9_21_1
e_1_2_9_46_1
e_1_2_9_23_1
e_1_2_9_44_1
e_1_2_9_7_1
e_1_2_9_5_1
e_1_2_9_3_1
e_1_2_9_9_1
e_1_2_9_25_1
e_1_2_9_27_1
e_1_2_9_48_1
e_1_2_9_29_1
References_xml – ident: e_1_2_9_45_1
  doi: 10.1093/nar/gky1047
– ident: e_1_2_9_12_1
  doi: 10.1093/nar/gkad376
– ident: e_1_2_9_29_1
  doi: 10.1021/bi952419l
– ident: e_1_2_9_20_1
– ident: e_1_2_9_35_1
  doi: 10.1038/s41586‐024‐07487‐w
– ident: e_1_2_9_46_1
  doi: 10.1093/nar/gkr703
– ident: e_1_2_9_32_1
  doi: 10.1021/jacs.2c12462
– ident: e_1_2_9_47_1
– ident: e_1_2_9_41_1
  doi: 10.1016/0263‐7855(90)80070‐V
– ident: e_1_2_9_39_1
  doi: 10.1093/nar/gkm216
– ident: e_1_2_9_11_1
  doi: 10.1093/bib/bbab308
– ident: e_1_2_9_4_1
  doi: 10.1111/j.1742-4658.2009.07360.x
– ident: e_1_2_9_33_1
  doi: 10.1002/prot.26593
– ident: e_1_2_9_48_1
  doi: 10.1093/nar/gkg571
– ident: e_1_2_9_3_1
  doi: 10.1111/j.1399-6576.1997.tb04627.x
– ident: e_1_2_9_14_1
  doi: 10.1016/j.str.2022.11.012
– ident: e_1_2_9_5_1
  doi: 10.3389/fmolb.2021.697586
– ident: e_1_2_9_23_1
  doi: 10.1016/j.toxicon.2018.05.006
– ident: e_1_2_9_17_1
  doi: 10.1101/2023.05.24.542179
– ident: e_1_2_9_50_1
  doi: 10.1093/bioinformatics/btt473
– ident: e_1_2_9_26_1
  doi: 10.1039/D2SC00924B
– ident: e_1_2_9_40_1
  doi: 10.1107/S0021889892009944
– ident: e_1_2_9_44_1
  doi: 10.1038/nature08142
– ident: e_1_2_9_52_1
  doi: 10.1073/pnas.1111471108
– ident: e_1_2_9_7_1
  doi: 10.1038/s41392‐022‐00904‐4
– ident: e_1_2_9_21_1
  doi: 10.1093/nar/28.1.235
– ident: e_1_2_9_25_1
  doi: 10.1002/anie.201501714
– ident: e_1_2_9_31_1
  doi: 10.1002/pep2.24073
– ident: e_1_2_9_30_1
  doi: 10.1016/S0006‐3495(03)74814‐0
– ident: e_1_2_9_51_1
  doi: 10.1107/S0907444913007051
– ident: e_1_2_9_38_1
  doi: 10.7554/eLife.75751
– ident: e_1_2_9_43_1
  doi: 10.1093/nar/gks1169
– ident: e_1_2_9_2_1
  doi: 10.1002/bip.20734
– ident: e_1_2_9_8_1
  doi: 10.1038/s41573‐020‐00135‐8
– ident: e_1_2_9_34_1
  doi: 10.1093/bib/bbae215
– ident: e_1_2_9_22_1
  doi: 10.1038/nsb1203-980
– ident: e_1_2_9_49_1
  doi: 10.1002/prot.26257
– ident: e_1_2_9_9_1
  doi: 10.1038/nsb1101-923
– ident: e_1_2_9_37_1
  doi: 10.1038/s41586‐023‐06832‐9
– ident: e_1_2_9_6_1
  doi: 10.1038/s41573‐021‐00337‐8
– ident: e_1_2_9_10_1
  doi: 10.1186/s13062-015-0103-4
– ident: e_1_2_9_19_1
– ident: e_1_2_9_13_1
  doi: 10.1002/prot.26237
– ident: e_1_2_9_16_1
  doi: 10.1038/s41586-021-03828-1
– ident: e_1_2_9_42_1
  doi: 10.1038/nbt.3988
– ident: e_1_2_9_15_1
  doi: 10.1038/s41586-021-03819-2
– ident: e_1_2_9_36_1
  doi: 10.1016/j.sbi.2024.102973
– ident: e_1_2_9_18_1
  doi: 10.1101/622803
– ident: e_1_2_9_28_1
  doi: 10.3390/toxins15100600
– ident: e_1_2_9_24_1
  doi: 10.1021/bi035412
– ident: e_1_2_9_27_1
  doi: 10.1002/(SICI)1097‐0282(199705)41:6<623::AID‐BIP3>3.0.CO;2‐S
SSID ssj0006936
Score 2.4837265
SecondaryResourceType online_first
Snippet While of primary importance in both the biomedical and therapeutic fields, peptides suffer from a relative lack of dedicated tools to predict efficiently and...
SourceID proquest
pubmed
crossref
SourceType Aggregation Database
Index Database
Title Comparative Analysis of Deep Learning‐Based Algorithms for Peptide Structure Prediction
URI https://www.ncbi.nlm.nih.gov/pubmed/41047732
https://www.proquest.com/docview/3257526497
WOSCitedRecordID wos001586797900001&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
journalDatabaseRights – providerCode: PRVWIB
  databaseName: Wiley Online Library Full Collection 2020
  customDbUrl:
  eissn: 1097-0134
  dateEnd: 99991231
  omitProxy: false
  ssIdentifier: ssj0006936
  issn: 0887-3585
  databaseCode: DRFUL
  dateStart: 19960101
  isFulltext: true
  titleUrlDefault: https://onlinelibrary.wiley.com
  providerName: Wiley-Blackwell
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwtV1db9MwFLW6DdheEGx8jI3JCMTLFGjz5eRx66gQqqYKNqh4iWzHLpW6pGqzafv33BvbaSo0CR54iRLHcaR7juxr3-tjQt6JNJIBk6Gnkyj1Qp70POHnodcVkQKHHMawekH_-5CdnyfjcTrqdLTbC3MzY0WR3N6m8_8KNZQB2Lh19h_gbhqFArgH0OEKsMP1r4Dvt_S825IjZ0rNnZzqxDuF0Ss_PplNysW0-mVUGY5HmOKSY3IhispiaGG0wEBOA571Ykco7mAXqpeuLj7gKFlni9icUDEtrTJr1cqq_8avUdzDfNKfmVh9OwPnJ5_Y8P4XxQsPneu6EiunTSND3BVRaHvA9mBWLlwDdg3Dr7VPuyaYrUy_20VR2J5Z1_yjVzcqsShc8QHV-NN2JTD-_KqGMkTVCWZXS9c1tN2rDbLlsyiFDnzr7OvgctiM2XGKQWwrXut_XP1qhzxyH697LvdMR2q35OIJeWznE_TE8OAp6ahil-wB7lV5dUff0zrDtw6d7JKHp-5uu-_O-dsjP1qEoY4wtNQUCUPXCUNXhKEAK7WEoQ1h6Iowz8jl4NNF_7Nnj9vwJPgolacDpXWcQD-cx1opX-YsjFSQC4mhYD-MOcx9eTeVUojI5-go5iph8OjLGFzF4DnZLMpCvSRUoeRlJFTMYx2GUiXoFzKdc5YEPSnYPnnrTJnNjapKZvSz_Qxtn9W23ydvnJUzsAlGsnihyutlFsBAE4Ern0JLL4z5m3YcXK_ufXNAdlYkPCSbYCL1mjyQN9V0uTgiG2ycHFmC_AaUgIB7
linkProvider Wiley-Blackwell
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Comparative+Analysis+of+Deep+Learning-Based+Algorithms+for+Peptide+Structure+Prediction&rft.jtitle=Proteins%2C+structure%2C+function%2C+and+bioinformatics&rft.au=Sauvestre%2C+Cl%C3%A9ment&rft.au=Zagury%2C+Jean-Fran%C3%A7ois&rft.au=Langenfeld%2C+Florent&rft.date=2025-10-05&rft.eissn=1097-0134&rft_id=info:doi/10.1002%2Fprot.70049&rft_id=info%3Apmid%2F41047732&rft.externalDocID=41047732
thumbnail_l http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=0887-3585&client=summon
thumbnail_m http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=0887-3585&client=summon
thumbnail_s http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=0887-3585&client=summon