Context-aware geometric deep learning for protein sequence design.

Protein design and engineering are evolving at an unprecedented pace leveraging the advances in deep learning. Current models nonetheless cannot natively consider non-protein entities within the design process. Here, we introduce a deep learning approach based solely on a geometric transformer of atomic coordinates and element names that predicts protein sequences from backbone scaffolds aware of the restraints imposed by diverse molecular environments. To validate the method, we show that it can produce highly thermostable, catalytically active enzymes with high success rates. This concept is anticipated to improve the versatility of protein design pipelines for crafting desired functions.

PeSTo-Carbs: Geometric Deep Learning for Prediction of Protein-Carbohydrate Binding Interfaces.

The Protein Structure Transformer (PeSTo), a geometric transformer, has exhibited exceptional performance in predicting protein-protein binding interfaces and distinguishing interfaces with nucleic acids, lipids, small molecules, and ions. In this study, we introduce PeSTo-Carbs, an extension of PeSTo specifically engineered to predict protein-carbohydrate binding interfaces. We evaluate the performance of this approach using independent test sets and compare them with those of previous methods. Furthermore, we highlight the model's capability to specialize in predicting interfaces involving cyclodextrins, a biologically and pharmaceutically significant class of carbohydrates. Our method consistently achieves remarkable accuracy despite the scarcity of available structural data for cyclodextrins.

Deep Learning-Assisted Single-Molecule Detection of Protein Post-translational Modifications with a Biological Nanopore.

Protein post-translational modifications (PTMs) play a crucial role in countless biological processes, profoundly modulating protein properties on both spatial and temporal scales. Protein PTMs have also emerged as reliable biomarkers for several diseases. However, only a handful of techniques are available to accurately measure their levels, capture their complexity at a single molecule level, and characterize their multifaceted roles in health and disease. Nanopore sensing provides high sensitivity for the detection of low-abundance proteins, holding the potential to impact single-molecule proteomics and PTM detection, in particular. Here, we demonstrate the ability of a biological nanopore, the pore-forming toxin aerolysin, to detect and distinguish α-synuclein-derived peptides bearing single or multiple PTMs, namely, phosphorylation, nitration, and oxidation occurring at different positions and in various combinations. The characteristic current signatures of the α-synuclein peptide and its PTM variants could be confidently identified by using a deep learning model for signal processing. We further demonstrate that this framework can quantify α-synuclein peptides at picomolar concentrations and detect the C-terminal peptides generated by digestion of full-length α-synuclein. Collectively, our work highlights the advantage of using nanopores as a tool for simultaneous detection of multiple PTMs and facilitates their use in biomarker discovery and diagnostics.

PeSTo: parameter-free geometric deep learning for accurate prediction of protein binding interfaces.

Proteins are essential molecular building blocks of life, responsible for most biological functions as a result of their specific molecular interactions. However, predicting their  binding  interfaces remains a challenge. In this study, we present a geometric transformer that acts directly on atomic coordinates labeled only with element names. The resulting model-the Protein Structure Transformer, PeSTo-surpasses the current state of the art in predicting protein-protein interfaces and can also predict and differentiate between interfaces involving nucleic acids, lipids, ions, and small molecules with high confidence. Its low computational cost enables processing high volumes of structural data, such as molecular dynamics ensembles allowing for the discovery of interfaces that remain otherwise inconspicuous in static experimentally solved structures. Moreover, the growing foldome provided by de novo structural predictions can be easily analyzed, providing new opportunities to uncover unexplored biology.

Visualization, Interactive Handling and Simulation of Molecules in Commodity Augmented Reality in Web Browsers Using moleculARweb's Virtual Modeling Kits.

moleculARweb (https://molecularweb.epfl.ch) began as a website for education and outreach in chemistry and structural biology through augmented reality (AR) content that runs in the web browsers of regular devices like smartphones, tablets, and computers. Here we present two evolutions of moleculARweb's Virtual Modeling Kits (VMK), tools where users can build and view molecules, and explore their mechanics, in 3D AR by handling the molecules in full 3D with custom-printed cube markers (VMK 2.0) or by moving around a simulated scene with mouse or touch gestures (VMK 3.0). Upon simulation the molecules experience visually realistic torsions, clashes, and hydrogen-bonding interactions that the user can manually switch on and off to explore their effects. Moreover, by manually tuning a fictitious temperature the users can accelerate conformational transitions or 'freeze' specific conformations for careful inspection in 3D. Even some phase transitions and separations can be simulated. We here showcase these and other features of the new VMKs connecting them to possible specific applications to teaching and self-learning of concepts from general, organic, biological and physical chemistry; and in assisting with small tasks in molecular modelling for research. Last, in a short discussion section we overview what future developments are needed for the 'dream tool' for the future of chemistry education and work.

Investigating Crosstalk Among PTMs Provides Novel Insight Into the Structural Basis Underlying the Differential Effects of Nt17 PTMs on Mutant Httex1 Aggregation.

Post-translational modifications (PTMs) within the first 17 amino acids (Nt17) of the Huntingtin protein (Htt) have been shown to inhibit the aggregation and attenuate the toxicity of mutant Htt proteins in vitro and in various models of Huntington's disease. Here, we expand on these studies by investigating the effect of methionine eight oxidation (oxM8) and its crosstalk with lysine 6 acetylation (AcK6) or threonine 3 phosphorylation (pT3) on the aggregation of mutant Httex1 (mHttex1). We show that M8 oxidation delays but does not inhibit the aggregation and has no effect on the final morphologies of mHttex1aggregates. The presence of both oxM8 and AcK6 resulted in dramatic inhibition of Httex1 fibrillization. Circular dichroism spectroscopy and molecular dynamics simulation studies show that PTMs that lower the mHttex1 aggregation rate (oxM8, AcK6/oxM8, pT3, pT3/oxM8, and pS13) result in increased population of a short N-terminal helix (first eight residues) in Nt17 or decreased abundance of other helical forms, including long helix and short C-terminal helix. PTMs that did not alter the aggregation rate (AcK6) of mHttex1 exhibit a similar distribution of helical conformation as the unmodified peptides. These results show that the relative abundance of N- vs. C-terminal helical conformations and long helices, rather than the overall helicity of Nt17, better explains the effect of different Nt17 PTMs on mHttex1; thus, explaining the lack of correlation between the effect of PTMs on the overall helicity of Nt17 and mHttex1 aggregation in vitro. Taken together, our results provide novel structural insight into the differential effects of single PTMs and crosstalk between different PTMs in regulating mHttex1 aggregation.

Aerolysin nanopores decode digital information stored in tailored macromolecular analytes.

Digital data storage is a growing need for our society and finding alternative solutions than those based on silicon or magnetic tapes is a challenge in the era of "big data." The recent development of polymers that can store information at the molecular level has opened up new opportunities for ultrahigh density data storage, long-term archival, anticounterfeiting systems, and molecular cryptography. However, synthetic informational polymers are so far only deciphered by tandem mass spectrometry. In comparison, nanopore technology can be faster, cheaper, nondestructive and provide detection at the single-molecule level; moreover, it can be massively parallelized and miniaturized in portable devices. Here, we demonstrate the ability of engineered aerolysin nanopores to accurately read, with single-bit resolution, the digital information encoded in tailored informational polymers alone and in mixed samples, without compromising information density. These findings open promising possibilities to develop writing-reading technologies to process digital data using a biological-inspired platform.

Analysis of the S. pombe Meiotic Proteome Reveals a Switch from Anabolic to Catabolic Processes and Extensive Post-transcriptional Regulation.

Meiotic progression in S. pombe is regulated by stage-specific gene expression and translation, changes in RNA stability, expression of anti-sense transcripts, and targeted proteolysis of regulatory proteins. We have used SILAC labeling to examine the relative levels of proteins in diploid S. pombe cells during meiosis. Among the 3,268 proteins quantified at all time points, the levels of 880 proteins changed at least 2-fold; the majority of proteins showed stepwise increases or decreases during the meiotic divisions, while some changed transiently. Overall, we observed reductions in proteins involved in anabolism and increases in proteins involved in catabolism. We also observed increases in the levels of proteins of the ESCRT-III complex and revealed a role for ESCRT-III components in chromosome segregation and spore formation. Correlation with studies of meiotic gene expression and ribosome occupancy reveals that many of the changes in steady-state protein levels are post-transcriptional.