CryoFold: Determining protein structures and data-guided ensembles from cryo-EM density maps

Mrinal Shekhar; Genki Terashi; Chitrak Gupta; Daipayan Sarkar; Gaspard Debussche; Nicholas J. Sisco; Jonathan Nguyen; Arup Mondal; John Vant; Petra Fromme; Wade D. Van Horn; Emad Tajkhorshid; Daisuke Kihara; Ken Dill; Alberto Perez; Abhishek Singharoy

doi:10.1016/j.matt.2021.09.004

CryoFold: Determining protein structures and data-guided ensembles from cryo-EM density maps

Mrinal Shekhar, Genki Terashi, Chitrak Gupta, Daipayan Sarkar, Gaspard Debussche, Nicholas J. Sisco, Jonathan Nguyen, Arup Mondal, John Vant, Petra Fromme, Wade D. Van Horn, Emad Tajkhorshid, Daisuke Kihara, Ken Dill, Alberto Perez, Abhishek Singharoy

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

27 Scopus citations

Abstract

Cryoelectron microscopy requires molecular modeling for refinement of structures. Ensemble models arrive at low free-energy molecular structures, but are computationally expensive and limited to resolving only small proteins. We introduce CryoFold, a pipeline of molecular dynamics simulations that determines ensembles of protein structures by integrating density data of varying sparsity at 3–5 Å resolution with sequence information and coarse-grained topological knowledge of the protein folds. We present six examples, folding proteins between 72 and 2,000 residues, including large membrane and multi-domain systems, and results from two Electron Microscopy Data Bank (EMDB) competitions. Driven by data from a single state, CryoFold discovers ensembles of common low-energy models together with rare low-probability structures that capture the equilibrium distribution of proteins constrained by the density maps. Many of these conformations are experimentally validated and functionally relevant. We arrive at a set of best practices for data-guided protein folding that are controlled using a Python graphical user interface (GUI).

Original language	English
Pages (from-to)	3195-3216
Number of pages	22
Journal	Matter
Volume	4
Issue number	10
DOIs	https://doi.org/10.1016/j.matt.2021.09.004
State	Published - Oct 6 2021
Externally published	Yes

Funding

A.S. and C.G. acknowledge start-up funds from the SMS and Biodesign Center for Applied Structure Discovery at Arizona State University , CAREER award by NSF -MCB 1942763 , and the resources of the OLCF at the Oak Ridge National Laboratory, which is supported by the Office of Science at DOE under contract no. DE-AC05-00OR22725 , made available via the INCITE program. The ET laboratory is supported by NIH ( P41-GM104601 ); E.T., A.S., and M.S. acknowledge NIH ( R01-GM067887 ). This research is part of the Blue Waters Sustained Petascale Computing project, which is supported by the National Science Foundation (awards OCI-0725070 and ACI-1238993 ) and the state of Illinois . K.D. and A.P. appreciate support from a PRAC computer allocation supported by NSF award ACI1514873 , support from NIH grant GM125813 , and the Laufer Center . A.P. appreciates start-up support from the University of Florida . D.K. acknowledges support from the NIH ( R01GM133840 and R01GM123055 ), the National Science Foundation ( MCB1925643 , DMS1614777 , CMMI1825941 , and DBI2003635 ), and the Purdue Institute of Drug Discovery . W.V.H. acknowledges NIH ( R01GM112077 ). A.S. and C.G. acknowledge start-up funds from the SMS and Biodesign Center for Applied Structure Discovery at Arizona State University, CAREER award by NSF-MCB 1942763, and the resources of the OLCF at the Oak Ridge National Laboratory, which is supported by the Office of Science at DOE under contract no. DE-AC05-00OR22725, made available via the INCITE program. The ET laboratory is supported by NIH (P41-GM104601); E.T. A.S. and M.S. acknowledge NIH (R01-GM067887). This research is part of the Blue Waters Sustained Petascale Computing project, which is supported by the National Science Foundation (awards OCI-0725070 and ACI-1238993) and the state of Illinois. K.D. and A.P. appreciate support from a PRAC computer allocation supported by NSF award ACI1514873, support from NIH grant GM125813, and the Laufer Center. A.P. appreciates start-up support from the University of Florida. D.K. acknowledges support from the NIH (R01GM133840 and R01GM123055), the National Science Foundation (MCB1925643, DMS1614777, CMMI1825941, and DBI2003635), and the Purdue Institute of Drug Discovery. W.V.H. acknowledges NIH (R01GM112077). M.S. and G.T. integrated the entire CryoFold pipeline, performed simulations and ROSETTA, EM validation, made figures, and wrote the manuscript. C.G. and G.D. designed the GUI. J.V. J.N. and A.M. designed the user guide and reported in the supplemental information. D.S. performed refinement of ATP synthase ensemble and performed refinements for the competition. N.J.S. performed de novo Rosetta validation. P.F. performed SFX experiments on flpp3 protein. W.D.V.H. performed Rosetta refinements and contributed to the supplemental information. E.T. contributed to the main text, and led the molecular dynamics simulation software (NAMD) developments. D.K. led MAINMAST developments. K.D. led MELD developments and wrote the manuscript. A.P. developed the program to combine MAINMAST, MELD, and MDFF, and wrote the manuscript. A.S. conceived the project, performed ReMDFF simulations, oversaw all the teams, and wrote the manuscript. The authors declare no competing interests.

Keywords

ATP synthase
CryoEM modeling
MAP3: Understanding
computations
cryoelectron microscopy
ensemble refinement
integrative modeling
molecular dynamics simulations
protein folding ensemble

Access to Document

10.1016/j.matt.2021.09.004

Cite this

Shekhar, M., Terashi, G., Gupta, C., Sarkar, D., Debussche, G., Sisco, N. J., Nguyen, J., Mondal, A., Vant, J., Fromme, P., Van Horn, W. D., Tajkhorshid, E., Kihara, D., Dill, K., Perez, A., & Singharoy, A. (2021). CryoFold: Determining protein structures and data-guided ensembles from cryo-EM density maps. Matter, 4(10), 3195-3216. https://doi.org/10.1016/j.matt.2021.09.004

@article{36784c6119594995bf78f92efa7c2c32,

title = "CryoFold: Determining protein structures and data-guided ensembles from cryo-EM density maps",

abstract = "Cryoelectron microscopy requires molecular modeling for refinement of structures. Ensemble models arrive at low free-energy molecular structures, but are computationally expensive and limited to resolving only small proteins. We introduce CryoFold, a pipeline of molecular dynamics simulations that determines ensembles of protein structures by integrating density data of varying sparsity at 3–5 {\AA} resolution with sequence information and coarse-grained topological knowledge of the protein folds. We present six examples, folding proteins between 72 and 2,000 residues, including large membrane and multi-domain systems, and results from two Electron Microscopy Data Bank (EMDB) competitions. Driven by data from a single state, CryoFold discovers ensembles of common low-energy models together with rare low-probability structures that capture the equilibrium distribution of proteins constrained by the density maps. Many of these conformations are experimentally validated and functionally relevant. We arrive at a set of best practices for data-guided protein folding that are controlled using a Python graphical user interface (GUI).",

keywords = "ATP synthase, CryoEM modeling, MAP3: Understanding, computations, cryoelectron microscopy, ensemble refinement, integrative modeling, molecular dynamics simulations, protein folding ensemble",

author = "Mrinal Shekhar and Genki Terashi and Chitrak Gupta and Daipayan Sarkar and Gaspard Debussche and Sisco, {Nicholas J.} and Jonathan Nguyen and Arup Mondal and John Vant and Petra Fromme and {Van Horn}, {Wade D.} and Emad Tajkhorshid and Daisuke Kihara and Ken Dill and Alberto Perez and Abhishek Singharoy",

note = "Publisher Copyright: {\textcopyright} 2021",

year = "2021",

month = oct,

day = "6",

doi = "10.1016/j.matt.2021.09.004",

language = "English",

volume = "4",

pages = "3195--3216",

journal = "Matter",

issn = "2590-2393",

publisher = "Cell Press",

number = "10",

}

TY - JOUR

T1 - CryoFold

T2 - Determining protein structures and data-guided ensembles from cryo-EM density maps

AU - Shekhar, Mrinal

AU - Terashi, Genki

AU - Gupta, Chitrak

AU - Sarkar, Daipayan

AU - Debussche, Gaspard

AU - Sisco, Nicholas J.

AU - Nguyen, Jonathan

AU - Mondal, Arup

AU - Vant, John

AU - Fromme, Petra

AU - Van Horn, Wade D.

AU - Tajkhorshid, Emad

AU - Kihara, Daisuke

AU - Dill, Ken

AU - Perez, Alberto

AU - Singharoy, Abhishek

PY - 2021/10/6

Y1 - 2021/10/6

N2 - Cryoelectron microscopy requires molecular modeling for refinement of structures. Ensemble models arrive at low free-energy molecular structures, but are computationally expensive and limited to resolving only small proteins. We introduce CryoFold, a pipeline of molecular dynamics simulations that determines ensembles of protein structures by integrating density data of varying sparsity at 3–5 Å resolution with sequence information and coarse-grained topological knowledge of the protein folds. We present six examples, folding proteins between 72 and 2,000 residues, including large membrane and multi-domain systems, and results from two Electron Microscopy Data Bank (EMDB) competitions. Driven by data from a single state, CryoFold discovers ensembles of common low-energy models together with rare low-probability structures that capture the equilibrium distribution of proteins constrained by the density maps. Many of these conformations are experimentally validated and functionally relevant. We arrive at a set of best practices for data-guided protein folding that are controlled using a Python graphical user interface (GUI).

AB - Cryoelectron microscopy requires molecular modeling for refinement of structures. Ensemble models arrive at low free-energy molecular structures, but are computationally expensive and limited to resolving only small proteins. We introduce CryoFold, a pipeline of molecular dynamics simulations that determines ensembles of protein structures by integrating density data of varying sparsity at 3–5 Å resolution with sequence information and coarse-grained topological knowledge of the protein folds. We present six examples, folding proteins between 72 and 2,000 residues, including large membrane and multi-domain systems, and results from two Electron Microscopy Data Bank (EMDB) competitions. Driven by data from a single state, CryoFold discovers ensembles of common low-energy models together with rare low-probability structures that capture the equilibrium distribution of proteins constrained by the density maps. Many of these conformations are experimentally validated and functionally relevant. We arrive at a set of best practices for data-guided protein folding that are controlled using a Python graphical user interface (GUI).

KW - ATP synthase

KW - CryoEM modeling

KW - MAP3: Understanding

KW - computations

KW - cryoelectron microscopy

KW - ensemble refinement

KW - integrative modeling

KW - molecular dynamics simulations

KW - protein folding ensemble

UR - http://www.scopus.com/inward/record.url?scp=85122986425&partnerID=8YFLogxK

U2 - 10.1016/j.matt.2021.09.004

DO - 10.1016/j.matt.2021.09.004

M3 - Article

AN - SCOPUS:85122986425

SN - 2590-2393

VL - 4

SP - 3195

EP - 3216

JO - Matter

JF - Matter

IS - 10

ER -

CryoFold: Determining protein structures and data-guided ensembles from cryo-EM density maps

Abstract

Funding

Keywords

Access to Document

Other files and links

Fingerprint

Cite this