3D Structure From 2D Microscopy Images Using Deep Learning
Understanding the structure of a protein complex is crucial in determining its function. However, retrieving accurate 3D structures from microscopy images is highly challenging, particularly as many imaging modalities are two-dimensional. Recent advances in Artificial Intelligence have been applied...
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Frontiers Media S.A.
2021-10-01
|
| Series: | Frontiers in Bioinformatics |
| Subjects: | |
| Online Access: | https://www.frontiersin.org/articles/10.3389/fbinf.2021.740342/full |
| _version_ | 1818691509607727104 |
|---|---|
| author | Benjamin Blundell Christian Sieben Suliana Manley Ed Rosten QueeLim Ch’ng Susan Cox |
| author_facet | Benjamin Blundell Christian Sieben Suliana Manley Ed Rosten QueeLim Ch’ng Susan Cox |
| author_sort | Benjamin Blundell |
| collection | DOAJ |
| description | Understanding the structure of a protein complex is crucial in determining its function. However, retrieving accurate 3D structures from microscopy images is highly challenging, particularly as many imaging modalities are two-dimensional. Recent advances in Artificial Intelligence have been applied to this problem, primarily using voxel based approaches to analyse sets of electron microscopy images. Here we present a deep learning solution for reconstructing the protein complexes from a number of 2D single molecule localization microscopy images, with the solution being completely unconstrained. Our convolutional neural network coupled with a differentiable renderer predicts pose and derives a single structure. After training, the network is discarded, with the output of this method being a structural model which fits the data-set. We demonstrate the performance of our system on two protein complexes: CEP152 (which comprises part of the proximal toroid of the centriole) and centrioles. |
| first_indexed | 2024-12-17T12:43:02Z |
| format | Article |
| id | doaj.art-3ac8918c9b5544deb4a522cd179b9b3c |
| institution | Directory Open Access Journal |
| issn | 2673-7647 |
| language | English |
| last_indexed | 2024-12-17T12:43:02Z |
| publishDate | 2021-10-01 |
| publisher | Frontiers Media S.A. |
| record_format | Article |
| series | Frontiers in Bioinformatics |
| spelling | doaj.art-3ac8918c9b5544deb4a522cd179b9b3c2022-12-21T21:47:52ZengFrontiers Media S.A.Frontiers in Bioinformatics2673-76472021-10-01110.3389/fbinf.2021.7403427403423D Structure From 2D Microscopy Images Using Deep LearningBenjamin Blundell0Christian Sieben1Suliana Manley2Ed Rosten3QueeLim Ch’ng4Susan Cox5Centre for Developmental Biology, Institute of Psychiatry, Psychology and Neuroscience, King’s College London, London, United KingdomNanoscale Infection Biology Lab (NIBI), Helmholtz Centre for Infection Research, London, GermanyÉcole Polytechnique Fédérale de Lausanne, Lausanne, SwitzerlandSnap, Inc., London, United KingdomCentre for Developmental Biology, Institute of Psychiatry, Psychology and Neuroscience, King’s College London, London, United KingdomRandall Centre for Cell and Molecular Biophysics, King’s College London, London, United KingdomUnderstanding the structure of a protein complex is crucial in determining its function. However, retrieving accurate 3D structures from microscopy images is highly challenging, particularly as many imaging modalities are two-dimensional. Recent advances in Artificial Intelligence have been applied to this problem, primarily using voxel based approaches to analyse sets of electron microscopy images. Here we present a deep learning solution for reconstructing the protein complexes from a number of 2D single molecule localization microscopy images, with the solution being completely unconstrained. Our convolutional neural network coupled with a differentiable renderer predicts pose and derives a single structure. After training, the network is discarded, with the output of this method being a structural model which fits the data-set. We demonstrate the performance of our system on two protein complexes: CEP152 (which comprises part of the proximal toroid of the centriole) and centrioles.https://www.frontiersin.org/articles/10.3389/fbinf.2021.740342/fullSMLMdeep-learningstructurestormAI |
| spellingShingle | Benjamin Blundell Christian Sieben Suliana Manley Ed Rosten QueeLim Ch’ng Susan Cox 3D Structure From 2D Microscopy Images Using Deep Learning Frontiers in Bioinformatics SMLM deep-learning structure storm AI |
| title | 3D Structure From 2D Microscopy Images Using Deep Learning |
| title_full | 3D Structure From 2D Microscopy Images Using Deep Learning |
| title_fullStr | 3D Structure From 2D Microscopy Images Using Deep Learning |
| title_full_unstemmed | 3D Structure From 2D Microscopy Images Using Deep Learning |
| title_short | 3D Structure From 2D Microscopy Images Using Deep Learning |
| title_sort | 3d structure from 2d microscopy images using deep learning |
| topic | SMLM deep-learning structure storm AI |
| url | https://www.frontiersin.org/articles/10.3389/fbinf.2021.740342/full |
| work_keys_str_mv | AT benjaminblundell 3dstructurefrom2dmicroscopyimagesusingdeeplearning AT christiansieben 3dstructurefrom2dmicroscopyimagesusingdeeplearning AT sulianamanley 3dstructurefrom2dmicroscopyimagesusingdeeplearning AT edrosten 3dstructurefrom2dmicroscopyimagesusingdeeplearning AT queelimchng 3dstructurefrom2dmicroscopyimagesusingdeeplearning AT susancox 3dstructurefrom2dmicroscopyimagesusingdeeplearning |