Residue-level error detection in cryoelectron microscopy models

Gabriella Reggiano; Wolfgang Lugmayr; Daniel Farrell; Thomas C Marlovits; Frank DiMaio

doi:10.1016/j.str.2023.05.002

Residue-level error detection in cryoelectron microscopy models

Standard

Residue-level error detection in cryoelectron microscopy models. / Reggiano, Gabriella; Lugmayr, Wolfgang; Farrell, Daniel; Marlovits, Thomas C; DiMaio, Frank.

In: STRUCTURE, Vol. 31, No. 7, 06.07.2023, p. 860-869.e4.

Research output: SCORING: Contribution to journal › SCORING: Journal article › Research › peer-review

Harvard

Reggiano, G, Lugmayr, W, Farrell, D, Marlovits, TC & DiMaio, F 2023, 'Residue-level error detection in cryoelectron microscopy models', STRUCTURE, vol. 31, no. 7, pp. 860-869.e4. https://doi.org/10.1016/j.str.2023.05.002

APA

Reggiano, G., Lugmayr, W., Farrell, D., Marlovits, T. C., & DiMaio, F. (2023). Residue-level error detection in cryoelectron microscopy models. STRUCTURE, 31(7), 860-869.e4. https://doi.org/10.1016/j.str.2023.05.002

Vancouver

Reggiano G, Lugmayr W, Farrell D, Marlovits TC, DiMaio F. Residue-level error detection in cryoelectron microscopy models. STRUCTURE. 2023 Jul 6;31(7):860-869.e4. https://doi.org/10.1016/j.str.2023.05.002

Bibtex

@article{cc6dbd5cf0124cdbba41d351b91130c4,

title = "Residue-level error detection in cryoelectron microscopy models",

abstract = "Building accurate protein models into moderate resolution (3-5 {\AA}) cryoelectron microscopy (cryo-EM) maps is challenging and error prone. We have developed MEDIC (Model Error Detection in Cryo-EM), a robust statistical model that identifies local backbone errors in protein structures built into cryo-EM maps by combining local fit-to-density with deep-learning-derived structural information. MEDIC is validated on a set of 28 structures that were subsequently solved to higher resolutions, where we identify the differences between low- and high-resolution structures with 68% precision and 60% recall. We additionally use this model to fix over 100 errors in 12 deposited structures and to identify errors in 4 refined AlphaFold predictions with 80% precision and 60% recall. As modelers more frequently use deep learning predictions as a starting point for refinement and rebuilding, MEDIC's ability to handle errors in structures derived from hand-building and machine learning methods makes it a powerful tool for structural biologists.",

author = "Gabriella Reggiano and Wolfgang Lugmayr and Daniel Farrell and Marlovits, {Thomas C} and Frank DiMaio",

note = "Copyright {\textcopyright} 2023. Published by Elsevier Ltd.",

year = "2023",

month = jul,

day = "6",

doi = "10.1016/j.str.2023.05.002",

language = "English",

volume = "31",

pages = "860--869.e4",

journal = "STRUCTURE",

issn = "0969-2126",

publisher = "Cell Press",

number = "7",

}

RIS

TY - JOUR

T1 - Residue-level error detection in cryoelectron microscopy models

AU - Reggiano, Gabriella

AU - Lugmayr, Wolfgang

AU - Farrell, Daniel

AU - Marlovits, Thomas C

AU - DiMaio, Frank

PY - 2023/7/6

Y1 - 2023/7/6

N2 - Building accurate protein models into moderate resolution (3-5 Å) cryoelectron microscopy (cryo-EM) maps is challenging and error prone. We have developed MEDIC (Model Error Detection in Cryo-EM), a robust statistical model that identifies local backbone errors in protein structures built into cryo-EM maps by combining local fit-to-density with deep-learning-derived structural information. MEDIC is validated on a set of 28 structures that were subsequently solved to higher resolutions, where we identify the differences between low- and high-resolution structures with 68% precision and 60% recall. We additionally use this model to fix over 100 errors in 12 deposited structures and to identify errors in 4 refined AlphaFold predictions with 80% precision and 60% recall. As modelers more frequently use deep learning predictions as a starting point for refinement and rebuilding, MEDIC's ability to handle errors in structures derived from hand-building and machine learning methods makes it a powerful tool for structural biologists.

AB - Building accurate protein models into moderate resolution (3-5 Å) cryoelectron microscopy (cryo-EM) maps is challenging and error prone. We have developed MEDIC (Model Error Detection in Cryo-EM), a robust statistical model that identifies local backbone errors in protein structures built into cryo-EM maps by combining local fit-to-density with deep-learning-derived structural information. MEDIC is validated on a set of 28 structures that were subsequently solved to higher resolutions, where we identify the differences between low- and high-resolution structures with 68% precision and 60% recall. We additionally use this model to fix over 100 errors in 12 deposited structures and to identify errors in 4 refined AlphaFold predictions with 80% precision and 60% recall. As modelers more frequently use deep learning predictions as a starting point for refinement and rebuilding, MEDIC's ability to handle errors in structures derived from hand-building and machine learning methods makes it a powerful tool for structural biologists.

U2 - 10.1016/j.str.2023.05.002

DO - 10.1016/j.str.2023.05.002

M3 - SCORING: Journal article

C2 - 37253357

VL - 31

SP - 860-869.e4

JO - STRUCTURE

JF - STRUCTURE

SN - 0969-2126

IS - 7

ER -