Spherical convolutions and their application in molecular modelling

By Wouter Boomsma, Jes Frellsen. 2017

95 citations recorded by the source corpus.

Subfields: Model Architectures, Graph Learning, Computational Chemistry & Drug Discovery

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Model Architectures

In Graph Learning

In Computational Chemistry & Drug Discovery

Citations in the corpus

Gauge Equivariant Convolutional Networks and the Icosahedral CNN (2019, methodology): “…we present in this paper a speciﬁc model (the Icosahedral CNN), which can be viewed as a fast and simple alternative to Spherical CNNs (Cohen et al., 2018b; Esteves et al., 2018; Boomsma & Frellsen, 2017; Su & Grauman, 2017; Perraudin et al., 2018; Jiang et al., 2018; Kondor et al., 2018a).”
Clebsch-Gordan Nets: a Fully Fourier Space Spherical Convolutional Neural Network (2018, background): “A number of authors have addressed this problem in different ways [14, 15].”
Spherical CNNs (2018, background): “Other very recent work is concerned with the analysis of spherical images, but does not deﬁne an equivariant architecture (Su and Grauman, 2017; Boomsma and Frellsen, 2017).”
ProGen: Language Modeling for Protein Generation (2020, background): “The aforementioned work, along with O’Connell et al. (2018), Boomsma & Frellsen (2017), and Greener et al. (2018), all utilize explicit structural information for generative modeling, thereby are unable to fully capture the number and diversity of sequence-only data available.”
Learning inverse folding from millions of predicted structures (2022, methodology): “…deep learning methods have also been applied to predict amino acid propensities given the surrounding local structural environments (Anand-Achim et al., 2021; Boomsma & Frellsen, 2017; Shroff et al., 2020; Li et al., 2020; Qi & Zhang, 2020; Zhang et al., 2020; Chen et al., 2019; Wang et al., 2018).”
Protein sequence design with a learned potential (2022, methodology): “For the amino-acid type prediction task, our conditional model achieves a 57.3% test set accuracy, either outperforming 40 or matching 26,41,42 previously reported machine learning models for the same task.”
Kernel Transformer Networks for Compact Spherical Convolution (2018, methodology): “Finally, other methods attempt to reduce distortion by unwrapping a single 360◦ image into multiple images using perspective projection locally [3,7], e.”
Protein sequence design with a learned potential (2020, methodology): “Previous machine learning models for this particular task have only been applied to architecture class, secondary structure, and ∆∆G prediction [66, 67, 68].”
Deep Learning 3D Shapes Using Alt-az Anisotropic 2-Sphere Convolution (2018, methodology): “Recently, researchers have started to explore deep spherical convolutional neural networks for tasks such as molecular modeling (Boomsma & Frellsen (2017)), ominidirectional vision (Su & Grauman (2017)) and 3D shape recognition (Cohen et al. (2018); Esteves et al. (2018)).”
Generative Models for Graph-Based Protein Design (2019, background): “Multiple works [16, 17] have modeled the conditional distribution of single amino acids given surrounding structure and sequence context with convolutional neural networks.”
DeepSphere: Efficient spherical Convolutional Neural Network with HEALPix sampling for cosmological applications (2018, methodology): “A ﬁrst approach, explored by [20] for molecular modeling and [21, 22] for omnidirectional imaging, is to use a 2D CNN on a discretisation of the sphere that is a grid, such as the equirectangular projection (Figure 6, left panel), or a set of grids, such as the cubed-sphere deﬁned by [47] (Figure…”
DeepSphere: a graph-based spherical CNN (2020, methodology): “Exploiting the locally Euclidean property of the sphere, early attempts used standard 2D convolutions on a grid sampling of the sphere (Boomsma & Frellsen, 2017; Su & Grauman, 2017; Coors et al., 2018).”
Molecular Property Prediction: A Multilevel Quantum Interactions Modeling Perspective (2019, methodology): “Along this line, convolutional neural network based models were proposed, where they represented each molecule as grid-like structures, such as image (Goh et al. 2017), string (G ´ omez-Bombarelli et al. 2018), and sphere (Boomsma and Frellsen 2017).”
Predicting changes in protein thermodynamic stability upon point mutation with deep 3D convolutional neural networks (2020, methodology): “Recently, a spherical convolutional architecture in which 474 concentric voxel grids parameterized by atom masses and charges were used as input to 475 predict the ∆∆Gs of direct mutations with good accuracy (49).”
Improving Data‐Driven Global Weather Prediction Using Deep Convolutional Neural Networks on a Cubed Sphere (2020, background): “produce saliency maps from 360 images and videos, demonstrating that it is able to propagate saliency maps across edges of the cube map and outperform CNNs applied to standard equirectangular grids. [3] apply a CNN with a cube map representation to classiﬁcation of three-dimensional molecular models”