Deep Learning for 3-D reconstruction of heterogeneous molecular structures from Cryo-EM data
Lead Research Organisation:
Imperial College London
Department Name: Electrical and Electronic Engineering
Abstract
Determining the structure of biomolecules is the goal of structural
biology and is essential to understand biological mechanisms
responsible for life and in drug discovery. Biological macromolecules
can be thought of as complex machines that perform complex
operations in living cells. These dynamic machines pass through
various conformations in the course of their actions and a complete
understanding of their working requires the determination of multiple
conformations.
Single particle Electron Cryo-Microscopy (Cryo-EM) has emerged as a unique method to determine molecular structures at near-atomic resolution. Achieving high-resolution estimation of structures of dynamic protein complexes requires large numbers of images and computationally intensive algorithms. Such reconstruction problems have been often approached by devising methods that use information about the imaging procedure and the properties of the object that needs to be reconstructed and many remarkable breakthroughs have been achieved over the years. These computational approaches are called "model-based" methods and
have the advantage to be predictable and stable. However, they can be computationally expensive and do not always derive maximum value from complex data. In particular, they are often unable to resolve complex heterogeneous structures. In contrast "data-driven" methods like deep neural networks have demonstrated, in other contexts, a remarkable ability to improve the quality of biomedical images. The problem with many deep learning approaches is that they are not predictable in the sense that often even small deviations in the input
data can result in a huge deviation of the output, which can have devastating effects in bio-imaging applications. Moreover, it is often very difficult to interpret what a deep network machine is really optimizing.
This project will advance a new family of deep neural networks for the
3-D reconstruction of dynamic protein complexes from cryo-electron
data. By working closely with structural biologists, we will put forward
approaches that systematically embed prior knowledge and constraints
about the signal and the physics of the data formation process into the
deep neural network architectures. We will also collaborate with Prof.
M. Unser and his team from EPFL Switzerland. They will provide
expertise in the area of analysis of stability of deep neural networks
and will share their experience in developing AI-based methods for
molecule reconstruction from Cryo-EM data. We expect that this
approach and these collaborations will allow us to introduce stable and
interpretable neural networks able to resolve heterogeneous biological
structures at a resolution that current methods cannot. The methods
produced will be in open-source format, integrated in existing
computational suites like CCP-EM and made available to the broadest
possible community.
biology and is essential to understand biological mechanisms
responsible for life and in drug discovery. Biological macromolecules
can be thought of as complex machines that perform complex
operations in living cells. These dynamic machines pass through
various conformations in the course of their actions and a complete
understanding of their working requires the determination of multiple
conformations.
Single particle Electron Cryo-Microscopy (Cryo-EM) has emerged as a unique method to determine molecular structures at near-atomic resolution. Achieving high-resolution estimation of structures of dynamic protein complexes requires large numbers of images and computationally intensive algorithms. Such reconstruction problems have been often approached by devising methods that use information about the imaging procedure and the properties of the object that needs to be reconstructed and many remarkable breakthroughs have been achieved over the years. These computational approaches are called "model-based" methods and
have the advantage to be predictable and stable. However, they can be computationally expensive and do not always derive maximum value from complex data. In particular, they are often unable to resolve complex heterogeneous structures. In contrast "data-driven" methods like deep neural networks have demonstrated, in other contexts, a remarkable ability to improve the quality of biomedical images. The problem with many deep learning approaches is that they are not predictable in the sense that often even small deviations in the input
data can result in a huge deviation of the output, which can have devastating effects in bio-imaging applications. Moreover, it is often very difficult to interpret what a deep network machine is really optimizing.
This project will advance a new family of deep neural networks for the
3-D reconstruction of dynamic protein complexes from cryo-electron
data. By working closely with structural biologists, we will put forward
approaches that systematically embed prior knowledge and constraints
about the signal and the physics of the data formation process into the
deep neural network architectures. We will also collaborate with Prof.
M. Unser and his team from EPFL Switzerland. They will provide
expertise in the area of analysis of stability of deep neural networks
and will share their experience in developing AI-based methods for
molecule reconstruction from Cryo-EM data. We expect that this
approach and these collaborations will allow us to introduce stable and
interpretable neural networks able to resolve heterogeneous biological
structures at a resolution that current methods cannot. The methods
produced will be in open-source format, integrated in existing
computational suites like CCP-EM and made available to the broadest
possible community.