Dataset Information

Deep learning the structural determinants of protein biochemical properties by comparing structural ensembles with DiffNets.

ABSTRACT: Understanding the structural determinants of a protein's biochemical properties, such as activity and stability, is a major challenge in biology and medicine. Comparing computer simulations of protein variants with different biochemical properties is an increasingly powerful means to drive progress. However, success often hinges on dimensionality reduction algorithms for simplifying the complex ensemble of structures each variant adopts. Unfortunately, common algorithms rely on potentially misleading assumptions about what structural features are important, such as emphasizing larger geometric changes over smaller ones. Here we present DiffNets, self-supervised autoencoders that avoid such assumptions, and automatically identify the relevant features, by requiring that the low-dimensional representations they learn are sufficient to predict the biochemical differences between protein variants. For example, DiffNets automatically identify subtle structural signatures that predict the relative stabilities of β-lactamase variants and duty ratios of myosin isoforms. DiffNets should also be applicable to understanding other perturbations, such as ligand binding.

SUBMITTER: Ward MD

PROVIDER: S-EPMC8140102 | biostudies-literature | 2021 May

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Deep learning the structural determinants of protein biochemical properties by comparing structural ensembles with DiffNets.

Ward Michael D MD Zimmerman Maxwell I MI Meller Artur A Chung Moses M Swamidass S J SJ Bowman Gregory R GR

Nature communications 20210521 1

Understanding the structural determinants of a protein's biochemical properties, such as activity and stability, is a major challenge in biology and medicine. Comparing computer simulations of protein variants with different biochemical properties is an increasingly powerful means to drive progress. However, success often hinges on dimensionality reduction algorithms for simplifying the complex ensemble of structures each variant adopts. Unfortunately, common algorithms rely on potentially misle ...[more]

PMID: 34021153

Similar Datasets

Project description:The plant cuticle, which consists of cutin and waxes, forms a hydrophobic coating covering the aerial surfaces of all plants. It acts as an interface between plants and their surrounding environment whilst also protecting them against biotic and abiotic stresses. In this research, we have investigated the biodiversity and cuticle properties of aquatic plant duckweed, using samples isolated from four different locations around Hongze lake in Jiangsu province, China. The samples were genotyped using two chloroplast markers and nuclear ribosomal DNA markers, which revealed them as ecotypes of the larger duckweed, Spirodela polyrhiza. Duckweed cuticle properties were investigated by compositional analysis using Gas Chromatography coupled with Mass Spectroscopy (GC-MS) Flame Ionization Detector (GC-FID), and ultrastructural observation by cryo-Scanning Electron Microscopy (cryo-SEM). Cuticle compositional analysis indicated that fatty acids and primary alcohols, the two typical constituents found in many land plant cuticle, are the major duckweed wax components. A large portion of the duckweed wax fraction is composed of phytosterols, represented by campesterol, stigmasterol, sitosterol and their common precursor squalene. The cryo-SEM observation uncovered significant differences between the surface structures of the top air-facing and bottom water-facing sides of the plant fronds. The top side of the fronds, containing multiple stomata complexes, appeared to be represented by a rather flat waxy film sporadically covered with wax crystals. Underneath the waxy film was detected a barely distinguished nanoridge net, which became distinctly noticeable after chloroform treatment. On the bottom side of the fronds, the large epidermal cells were covered by the well-structured net, whose sections became narrower and sharper under cryo-SEM following chloroform treatment. These structural differences between the abaxial and adaxial sides of the fronds evidently relate to their distinct physiological roles in interacting with the contrasting environments of sunlight/air and nutrients/water. The unique structural and biochemical features of Spirodela frond surfaces with their rapid reproductive cycle and readily availability genome sequence, make duckweed an attractive monocot model for studying the fundamental processes related to plant protection against ultraviolet irradiation, pathogens and other environmental stresses.

Dataset Information

Deep learning the structural determinants of protein biochemical properties by comparing structural ensembles with DiffNets.

Publications

Deep learning the structural determinants of protein biochemical properties by comparing structural ensembles with DiffNets.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets