Latent generative landscapes as maps of functional diversity in protein sequence space

Variational autoencoders are unsupervised learning models with generative capabilities, when applied to protein data, they classify sequences by phylogeny and generate de novo sequences which preserve statistical properties of protein composition. While previous studies focus on clustering and gener...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Nature communications Jg. 14; H. 1; S. 2222 - 15
Hauptverfasser: Ziegler, Cheyenne, Martin, Jonathan, Sinner, Claude, Morcos, Faruck
Format: Journal Article
Sprache:Englisch
Veröffentlicht: London Nature Publishing Group UK 19.04.2023
Nature Publishing Group
Nature Portfolio
Schlagworte:
631/114/1305
631/114/2404
631/114/469
631/57/2266
Amino Acid Sequence
beta-Lactamases - genetics
Clustering
Fitness
Globins
Hamiltonian functions
Humanities and Social Sciences
Ion channels
Manifolds
multidisciplinary
Phylogenetics
Phylogeny
Protein composition
Protein engineering
Protein families
Proteins
Reproductive fitness
Science
Science (multidisciplinary)
Transcription Factors
Unsupervised learning
β Lactamase
ISSN:2041-1723, 2041-1723
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Variational autoencoders are unsupervised learning models with generative capabilities, when applied to protein data, they classify sequences by phylogeny and generate de novo sequences which preserve statistical properties of protein composition. While previous studies focus on clustering and generative features, here, we evaluate the underlying latent manifold in which sequence information is embedded. To investigate properties of the latent manifold, we utilize direct coupling analysis and a Potts Hamiltonian model to construct a latent generative landscape. We showcase how this landscape captures phylogenetic groupings, functional and fitness properties of several systems including Globins, β -lactamases, ion channels, and transcription factors. We provide support on how the landscape helps us understand the effects of sequence variability observed in experimental data and provides insights on directed and natural protein evolution. We propose that combining generative properties and functional predictive power of variational autoencoders and coevolutionary analysis could be beneficial in applications for protein engineering and design. In this work, the authors study protein families’ VAE latent manifolds and coevolutionary Hamiltonians. These Latent Generative Landscapes predict phylogenetic groupings, fitness & functional properties for several systems with clear protein engineering/design potential.
Bibliographie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
ISSN:2041-1723
2041-1723
DOI:10.1038/s41467-023-37958-z