DeepAlign, a 3D Alignment Method based on Regionalized Deep
Learning for Cryo-EM

J 2021

DeepAlign, a 3D Alignment Method based on Regionalized Deep Learning for Cryo-EM

JIMÉNEZ-MORENO, Amaya, David STŘELÁK, Jiří FILIPOVIČ, José María CARAZO, Carlos Óscar S. SORZANO et. al.

Základní údaje

Originální název

DeepAlign, a 3D Alignment Method based on Regionalized Deep Learning for Cryo-EM

Autoři

JIMÉNEZ-MORENO, Amaya (724 Španělsko), David STŘELÁK (203 Česká republika, domácí), Jiří FILIPOVIČ (203 Česká republika, garant, domácí), José María CARAZO a Carlos Óscar S. SORZANO

Vydání

Journal of Structural Biology, San Diego,USA, Academic Press, 2021, 1047-8477

Další údaje

Jazyk

angličtina

Typ výsledku

Článek v odborném periodiku

Obor

10201 Computer sciences, information science, bioinformatics

Stát vydavatele

Spojené státy

Utajení

není předmětem státního či obchodního tajemství

Odkazy

URL

Impakt faktor

Impact factor: 3.234

Kód RIV

RIV/00216224:14610/21:00121351

Organizační jednotka

Ústav výpočetní techniky

DOI

http://dx.doi.org/10.1016/j.jsb.2021.107712

UT WoS

000756475200010

Klíčová slova anglicky

3D alignment; 3D reconstruction; Cryo-EM; Deep learning; Machine learning

Štítky

rivok

Příznaky

Mezinárodní význam, Recenzováno

Změněno: 12. 4. 2022 16:50, Mgr. Alena Mokrá

Anotace

V originále

Cryo Electron Microscopy (Cryo-EM) is currently one of the main tools to reveal the structural information of biological specimens at high resolution. Despite the great development of the techniques involved to solve the biological structures with Cryo-EM in the last years, the reconstructed 3D maps can present lower resolution due to errors committed while processing the information acquired by the microscope. One of the main problems comes from the 3D alignment step, which is an error-prone part of the reconstruction workflow due to the very low signal-to-noise ratio (SNR) common in Cryo-EM imaging. In fact, as we will show in this work, it is not unusual to find a disagreement in the alignment parameters in approximately 20–40% of the processed images, when outputs of different alignment algorithms are compared. In this work, we present a novel method to align sets of single particle images in the 3D space, called DeepAlign. Our proposal is based on deep learning networks that have been successfully used in plenty of problems in image classification. Specifically, we propose to design several deep neural networks on a regionalized basis to classify the particle images in sub-regions and, then, make a refinement of the 3D alignment parameters only inside that sub-region. We show that this method results in accurately aligned images, improving the Fourier shell correlation (FSC) resolution obtained with other state-of-the-art methods while decreasing computational time.

Návaznosti

MUNI/A/1411/2019, interní kód MU

Název: Aplikovaný výzkum: softwarové architektury kritických infrastruktur, bezpečnost počítačových systémů, zpracování přirozeného jazyka a jazykové inženýrství, vizualizaci velkých dat a rozšířená realita.

Investor: Masarykova univerzita, Aplikovaný výzkum: softwarové architektury kritických infrastruktur, bezpečnost počítačových systémů, zpracování přirozeného jazyka a jazykové inženýrství, vizualizaci velkých dat a rozšířená realita., DO R. 2020_Kategorie A - Specifický výzkum - Studentské výzkumné projekty

MUNI/A/1549/2020, interní kód MU

Název: Zapojení studentů Fakulty informatiky do mezinárodní vědecké komunity 21 (Akronym: SKOMU)

Investor: Masarykova univerzita, Zapojení studentů Fakulty informatiky do mezinárodní vědecké komunity 21

Citovat

JIMÉNEZ-MORENO, Amaya, David STŘELÁK, Jiří FILIPOVIČ, José María CARAZO a Carlos Óscar S. SORZANO. DeepAlign, a 3D Alignment Method based on Regionalized Deep Learning for Cryo-EM. Journal of Structural Biology. San Diego,USA: Academic Press, 2021, roč. 213, č. 2, s. "107712", 14 s. ISSN 1047-8477. Dostupné z: https://dx.doi.org/10.1016/j.jsb.2021.107712.

@article{1757923,
   author = {JiménezandMoreno, Amaya and Střelák, David and Filipovič, Jiří and Carazo, José María and Sorzano, Carlos Óscar S.},
   article_location = {San Diego,USA},
   article_number = {2},
   doi = {http://dx.doi.org/10.1016/j.jsb.2021.107712},
   keywords = {3D alignment; 3D reconstruction; Cryo-EM; Deep learning; Machine learning},
   language = {eng},
   issn = {1047-8477},
   journal = {Journal of Structural Biology},
   title = {DeepAlign, a 3D Alignment Method based on Regionalized Deep Learning for Cryo-EM},
   url = {https://www.sciencedirect.com/science/article/pii/S1047847721000174},
   volume = {213},
   year = {2021}
}

TY  - JOUR
ID  - 1757923
AU  - Jiménez-Moreno, Amaya - Střelák, David - Filipovič, Jiří - Carazo, José María - Sorzano, Carlos Óscar S.
PY  - 2021
TI  - DeepAlign, a 3D Alignment Method based on Regionalized Deep Learning for Cryo-EM
JF  - Journal of Structural Biology
VL  - 213
IS  - 2
SP  - "107712"
EP  - "107712"
PB  - Academic Press
SN  - 10478477
KW  - 3D alignment
KW  - 3D reconstruction
KW  - Cryo-EM
KW  - Deep learning
KW  - Machine learning
UR  - https://www.sciencedirect.com/science/article/pii/S1047847721000174
N2  - Cryo Electron Microscopy (Cryo-EM) is currently one of the main tools to reveal the structural information of biological specimens at high resolution. Despite the great development of the techniques involved to solve the biological structures with Cryo-EM in the last years, the reconstructed 3D maps can present lower resolution due to errors committed while processing the information acquired by the microscope. One of the main problems comes from the 3D alignment step, which is an error-prone part of the reconstruction workflow due to the very low signal-to-noise ratio (SNR) common in Cryo-EM imaging. In fact, as we will show in this work, it is not unusual to find a disagreement in the alignment parameters in approximately 20–40% of the processed images, when outputs of different alignment algorithms are compared. In this work, we present a novel method to align sets of single particle images in the 3D space, called DeepAlign. Our proposal is based on deep learning networks that have been successfully used in plenty of problems in image classification. Specifically, we propose to design several deep neural networks on a regionalized basis to classify the particle images in sub-regions and, then, make a refinement of the 3D alignment parameters only inside that sub-region. We show that this method results in accurately aligned images, improving the Fourier shell correlation (FSC) resolution obtained with other state-of-the-art methods while decreasing computational time.
ER  -

JIMÉNEZ-MORENO, Amaya, David STŘELÁK, Jiří FILIPOVIČ, José María CARAZO a Carlos Óscar S. SORZANO. DeepAlign, a 3D Alignment Method based on Regionalized Deep Learning for Cryo-EM. \textit{Journal of Structural Biology}. San Diego,USA: Academic Press, 2021, roč.~213, č.~2, s.~''107712'', 14 s. ISSN~1047-8477. Dostupné z: https://dx.doi.org/10.1016/j.jsb.2021.107712.

Podrobný výpis o publikaci