Protein function predictions

  • Article
    | Open Access

    Finding a biologically-relevant inductive bias for training DNNs on large fitness landscapes is challenging. Here, the authors propose a method called Epistatic Net that improves DNN prediction accuracy and interpretation speed by integrating the knowledge that higher-order epistatic interactions are usually sparse.

    • Amirali Aghazadeh
    • , Hunter Nisonoff
    •  & Kannan Ramchandran
  • Article
    | Open Access

    G protein-coupled receptors (GPCRs) are a critical target in modern drug development across a wide range of indications. Here the authors provide a comprehensive characterization of a typical GPCR, the angiotensin II (AngII) type 1 receptor (AT1R), and provide insight into its activation mechanism that suggest avenues for the design of allosteric GPCR modulators.

    • Shaoyong Lu
    • , Xinheng He
    •  & Jian Zhang
  • Article
    | Open Access

    The authors present flDPnn, a computational tool for disorder and disorder function predictions from protein sequences. flDPnn was assessed with the data from the “Critical Assessment of Protein Intrinsic Disorder Prediction” experiment and on an independent and low-similarity test dataset, which show that flDPnn offers accurate predictions of disorder, fully disordered proteins and four common disorder functions.

    • Gang Hu
    • , Akila Katuwawala
    •  & Lukasz Kurgan
  • Article
    | Open Access

    The rapid increase in the number of proteins in sequence databases and the diversity of their functions challenge computational approaches for automated function prediction. Here, the authors introduce DeepFRI, a Graph Convolutional Network for predicting protein functions by leveraging sequence features extracted from a protein language model and protein structures.

    • Vladimir Gligorijević
    • , P. Douglas Renfrew
    •  & Richard Bonneau
  • Article
    | Open Access

    The ability to design functional sequences is central to protein engineering and biotherapeutics. Here the authors introduce a deep generative alignment-free model for sequence design applied to highly variable regions and design and test a diverse nanobody library with improved properties for selection experiments.

    • Jung-Eun Shin
    • , Adam J. Riesselman
    •  & Debora S. Marks
  • Article
    | Open Access

    Our understanding of the residue-level details of protein interactions remains incomplete. Here, the authors show sequence coevolution can be used to infer interacting proteins with residue-level details, including predicting 467 interactions de novo in the Escherichia coli cell envelope proteome.

    • Anna G. Green
    • , Hadeer Elhabashy
    •  & Debora S. Marks
  • Article
    | Open Access

    Identifying variants capable of causing genetic disease is challenging. The authors use semisupervised learning to predict pathogenic missense variants and their impacts on protein structure and function, enabling a molecular mechanism-driven approach to studying different types of human disease.

    • Vikas Pejaver
    • , Jorge Urresti
    •  & Predrag Radivojac
  • Article
    | Open Access

    Protein-ligand unbinding processes are out of reach for atomistic simulations due to time-scale involved. Here the authors demonstrate an approach relying on dissipation-corrected targeted molecular dynamics that enables to provide binding and unbinding rates with a speed-up of several orders of magnitude.

    • Steffen Wolf
    • , Benjamin Lickert
    •  & Gerhard Stock
  • Article
    | Open Access

    A genetic diagnosis remains unattainable for many individuals with a rare disease because of incomplete knowledge about the genetic basis of many diseases. Here, the authors present the web-based tool GADO (GeneNetwork Assisted Diagnostic Optimization) that uses public RNA-seq data for prioritization of candidate genes.

    • Patrick Deelen
    • , Sipko van Dam
    •  & Lude Franke
  • Article
    | Open Access

    Allostery is a fundamental principle of protein regulation that remains challenging to engineer. Here authors screen human Inward Rectifier K + Channel Kir2.1 for permissibility to domain insertions and propose that differential permissibility is a metric of latent allosteric capacity in Kir2.1.

    • Willow Coyote-Maestas
    • , Yungui He
    •  & Daniel Schmidt
  • Article
    | Open Access

    Chordoid glioma is a slow growing diencephalic tumor whose mutational landscape is poorly characterized. Here, the authors perform whole-exome and RNA-sequencing and find that 15 of 16 chordoid glioma cases studied harbor the same PRKCA mutation which results in enhanced proliferation.

    • Shai Rosenberg
    • , Iva Simeonova
    •  & Marc Sanson
  • Article
    | Open Access

    So far no enzymatic activity has been attributed to flagellin, the major component of bacterial flagella. Here the authors use bioinformatic analysis and identify a metallopeptidase insertion in flagellins from 74 bacterial species and show that recombinant flagellin and flagellar filaments have proteolytic activity.

    • Ulrich Eckhard
    • , Hina Bandukwala
    •  & Andrew C. Doxey
  • Article
    | Open Access

    Protein stability modulation by E3 ubiquitin ligases is an important layer of functional regulation, but screening for E3 ligase-substrate interactions is time-consuming and costly. Here, the authors take an in silico naïve Bayesian classifier approach to integrate multiple lines of evidence for E3-substrate prediction, enabling prediction of the proteome-wide human E3 ligase interaction network.

    • Yang Li
    • , Ping Xie
    •  & Fuchu He
  • Article
    | Open Access

    Plastoquinone (PLQ) shuttles electrons between photosystem II (PSII) and cytochrome b6f. Here the authors perform molecular dynamics simulations and propose that PLQ enters the exchange cavity of PSII by a promiscuous diffusion mechanism whereby three different channels each act as entry and exit points.

    • Floris J. Van Eerden
    • , Manuel N. Melo
    •  & Siewert J. Marrink
  • Article
    | Open Access

    Proteins are sometimes implicated in separate and seemingly unrelated processes, so called moonlighting functions. Here the authors use bioinformatics tools to identify extreme multifunctional proteins and define a signature of extreme multifunctionality.

    • Charles E. Chapple
    • , Benoit Robisson
    •  & Christine Brun