Biology subjects

Baker, D.

Publications and source records attributed to Baker, D..

At least 217 records · Page 12Linked to original sources

bioRxiv2021

Interpreting Neural Networks for Biological Sequences by Learning Stochastic Masks

Sequence-based neural networks can learn to make accurate predictions from large biological datasets, but model interpretation remains challenging. Many existing feature attribution methods are optimized for continuous rather than discrete input patterns and assess individual feature importance in isolation, making them ill-suited for interpreting non-linear interactions in molecular sequences. Building on work in computer vision and natural language processing, we developed an approach based on deep generative modeling - Scrambler networks - wherein the most salient sequence positions are identified with learned input masks. Scramblers learn to generate Position-Specific Scoring Matrices (PSSMs) where unimportant nucleotides or residues are scrambled by raising their entropy. We apply Scramblers to interpret the effects of genetic variants, uncover non-linear interactions between cis-regulatory elements, explain binding specificity for protein-protein interactions, and identify structural determinants of de novo designed proteins. We show that interpretation based on a generative model allows for efficient attribution across large datasets and results in high-quality explanations, often outperforming state-of-the-art methods.

genomics↗

bioRxiv2021

Sentinel cells enable genetic detection of SARS-CoV-2 Spike protein

The COVID-19 pandemic has demonstrated the need for exploring different diagnostic and therapeutic modalities to tackle future viral threats. In this vein, we propose the idea of sentinel cells, cellular biosensors capable of detecting viral antigens and responding to them with customizable responses. Using SARS-CoV-2 as a test case, we developed a live cell sensor (SARSNotch) using a de novo-designed protein binder against the SARS-CoV-2 Spike protein. SARSNotch is capable of driving custom genetically-encoded payloads in immortalized cell lines or in primary T lymphocytes in response to purified SARS-CoV-2 Spike or in the presence of Spike-expressing cells. Furthermore, SARSNotch is functional in a cellular system used in directed evolution platforms for development of better binders or therapeutics. In keeping with the rapid dissemination of scientific knowledge that has characterized the incredible scientific response to the ongoing pandemic, we extend an open invitation for others to make use of and improve SARSNotch sentinel cells in the hopes of unlocking the potential of the next generation of smart antiviral therapeutics.

synthetic biology↗

bioRxiv2021

Ensuring scientific reproducibility in bio-macromolecular modeling via extensive, automated benchmarks

Each year vast international resources are wasted on irreproducible research. The scientific community has been slow to adopt standard software engineering practices, despite the increases in high-dimensional data, complexities of workflows, and computational environments. Here we show how scientific software applications can be created in a reproducible manner when simple design goals for reproducibility are met. We describe the implementation of a test server framework and 40 scientific benchmarks, covering numerous applications in Rosetta bio-macromolecular modeling. High performance computing cluster integration allows these benchmarks to run continuously and automatically. Detailed protocol captures are useful for developers and users of Rosetta and other macromolecular modeling tools. The framework and design concepts presented here are valuable for developers and users of any type of scientific software and for the scientific community to create reproducible methods. Specific examples highlight the utility of this framework and the comprehensive documentation illustrates the ease of adding new tests in a matter of hours.

bioinformatics↗

bioRxiv2021

A covariation analysis reveals elements of selectivity in quorum sensing systems

Many bacteria communicate with kin and coordinate group behaviors through a form of cell-cell signaling called acyl-homoserine lactone (AHL) quorum sensing (QS). In these systems, a signal synthase produces an AHL to which its paired receptor selectively responds. Selectivity is fundamental to cell signaling. Despite its importance, it has been challenging to determine how this selectivity is achieved and how AHL QS systems evolve and diversify. We hypothesized that we could use covariation within the protein sequences of AHL synthases and receptors to identify selectivity residues. We began by identifying about 6,000 unique synthase-receptor pairs. We then used the protein sequences of these pairs to identify covariation patterns and mapped the patterns onto the LasI/R system from Pseudomonas aeruginosa PAO1. The covarying residues in both proteins cluster around the ligand binding sites. We demonstrate that these residues are involved in system selectivity toward the cognate signal and go on to engineer the Las system to both produce and respond to an alternate AHL signal. We have thus demonstrated a new application for covariation methods and have deepened our understanding of how communication systems evolve and diversify.

microbiology↗

bioRxiv2021

Freely accessible ready to use global infrastructure for SARS-CoV-2 monitoring

The COVID-19 pandemic is the first global health crisis to occur in the age of big genomic data.Although data generation capacity is well established and sufficiently standardized, analytical capacity is not. To establish analytical capacity it is necessary to pull together global computational resources and deliver the best open source tools and analysis workflows within a ready to use, universally accessible resource. Such a resource should not be controlled by a single research group, institution, or country. Instead it should be maintained by a community of users and developers who ensure that the system remains operational and populated with current tools. A community is also essential for facilitating the types of discourse needed to establish best analytical practices. Bringing together public computational research infrastructure from the USA, Europe, and Australia, we developed a distributed data analysis platform that accomplishes these goals. It is immediately accessible to anyone in the world and is designed for the analysis of rapidly growing collections of deep sequencing datasets. We demonstrate its utility by detecting allelic variants in high-quality existing SARS-CoV-2 sequencing datasets and by continuous reanalysis of COG-UK data. All workflows, data, and documentation is available at https://covid19.galaxyproject.org.

bioinformatics↗

bioRxiv2021

Fast and accurate genome-wide predictions and structural modeling of protein-protein interactions using Galaxy

Protein-protein interactions play a crucial role in almost all cellular processes. Identifying interacting proteins reveals insight into living organisms and yields novel drug targets for disease treatment. Here, we present a publicly available, automated pipeline to predict genome-wide protein-protein interactions and produce high-quality multimeric structural models. Application of our method to the Human and Yeast genomes yield protein-protein interaction networks similar in quality to common experimental methods. We identified and modeled Human proteins likely to interact with the papain-like protease of SARS-CoV2s non-structural protein 3 (Nsp3). We also produced models of SARS-CoV2s spike protein (S) interacting with myelin-oligodendrocyte glycoprotein receptor (MOG) and dipeptidyl peptidase-4 (DPP4). The presented method is capable of confidently identifying interactions while providing high-quality multimeric structural models for experimental validation. The interactome modeling pipeline is available at usegalaxy.org and usegalaxy.eu.

genomics↗

bioRxiv2021

Large-scale design and refinement of stable proteins using sequence-only models

Engineered proteins generally must possess a stable structure in order to achieve their designed function. Stable designs, however, are astronomically rare within the space of all possible amino acid sequences. As a consequence, many designs must be tested computationally and experimentally in order to find stable ones, which is expensive in terms of time and resources. Here we report a neural network model that predicts protein stability based only on sequences of amino acids, and demonstrate its performance by evaluating the stability of almost 200,000 novel proteins. These include a wide range of sequence perturbations, providing a baseline for future work in the field. We also report a second neural network model that is able to generate novel stable proteins. Finally, we show that the predictive model can be used to substantially increase the stability of both expert-designed and model-generated proteins.

biophysics↗

bioRxiv2021

Sampling of Structure and Sequence Space of Small Protein Folds

Nature only samples a small fraction in sequence space, yet many more amino acid combinations can fold into stable proteins. Furthermore, small structural variations in a single fold, which may only be a few amino acids different from the next homolog, define their molecular function. Hence, to design proteins with novel molecular functionalities, such as molecular recognition, methods to control and sample shape diversity are necessary. To explore this space, we developed and experimentally validated a computational platform that can design a wide variety of small protein folds while sampling high shape diversity. We designed and evaluated about 30,000 de novo protein designs of 7 different folds. Among these designs, about 6,200 stable proteins were identified, with predicted structures having first-of-its-kind minimalized thioredoxin. Obtained data revealed more protein folding rules, such as helix connecting loops, which were in nature. Beyond providing a resource database for protein engineering, our data presents a large training data set for machine learning. We developed a high-accuracy classifier to predict the stability of our designed proteins. The methods and the wide range of new protein shapes provide a basis for the design of new protein function without compromising stability.

bioengineering↗

bioRxiv2021

Ultrapotent miniproteins targeting the receptor-binding domain protect against SARS-CoV-2 infection and disease in mice

Despite the introduction of public health measures and spike protein-based vaccines to mitigate the COVID-19 pandemic, SARS-CoV-2 infections and deaths continue to rise. Previously, we used a structural design approach to develop picomolar range miniproteins targeting the SARS-CoV-2 receptor binding domain. Here, we investigated the capacity of modified versions of one lead binder, LCB1, to protect against SARS-CoV-2-mediated lung disease in human ACE2-expressing transgenic mice. Systemic administration of LCB1-Fc reduced viral burden, diminished immune cell infiltration and inflammation, and completely prevented lung disease and pathology. A single intranasal dose of LCB1v1.3 reduced SARS-CoV-2 infection in the lung even when given as many as five days before or two days after virus inoculation. Importantly, LCB1v1.3 protected in vivo against a historical strain (WA1/2020), an emerging B.1.1.7 strain, and a strain encoding key E484K and N501Y spike protein substitutions. These data support development of LCB1v1.3 for prevention or treatment of SARS-CoV-2 infection.

microbiology↗

bioRxiv2021

Polyclonal antibody responses to HIV Env immunogens resolved using cryoEM

In BriefHerein, we evaluated the immunogenicity of several BG505 SOSIP-based HIV Env immunogens in the rhesus macaque animal model using a combination of serology and biophysical approaches. We applied electron cryo-microscopy for high-resolution mapping of elicited polyclonal antibody responses, which provided detailed insights into the binding modes of the most common classes of antibodies elicited by BG505 SOSIP immunogens as well as the critical differences in immunogenicity that can occur as a consequence of engineered stabilizing mutations and partial glycan occupancy at different sites. SummaryEngineered ectodomain trimer immunogens based on BG505 envelope glycoprotein are widely utilized as components of HIV vaccine development platforms. In this study, we used rhesus macaques to evaluate the immunogenicity of several stabilized BG505 SOSIP constructs both as free trimers and presented on a nanoparticle. We applied a cryoEM-based method for high-resolution mapping of polyclonal antibody responses elicited in immunized animals (cryoEMPEM). Mutational analysis coupled with neutralization assays were used to probe the neutralization potential at each epitope. We demonstrate that cryoEMPEM data can be used for rapid, high-resolution analysis of polyclonal antibody responses without the need for monoclonal antibody isolation. This approach allowed to resolve structurally distinct classes of antibodies that bind overlapping sites. In addition to comprehensive mapping of commonly targeted neutralizing and non-neutralizing epitopes in BG505 SOSIP immunogens, our analysis revealed that epitopes comprising engineered stabilizing mutations and of partially occupied glycosylation sites can be immunogenic. Graphical abstract O_FIG_DISPLAY_L [Figure 1] M_FIG_DISPLAY C_FIG_DISPLAY

immunology↗

bioRxiv2020

Single Layers of Attention Suffice to Predict Protein Contacts

AO_SCPLOWBSTRACTC_SCPLOWThe established approach to unsupervised protein contact prediction estimates co-evolving positions using undirected graphical models. This approach trains a Potts model on a Multiple Sequence Alignment, then predicts that the edges with highest weight correspond to contacts in the 3D structure. On the other hand, increasingly large Transformers are being pretrained on protein sequence databases but have demonstrated mixed results for downstream tasks, including contact prediction. This has sparked discussion about the role of scale and attention-based models in unsupervised protein representation learning. We argue that attention is a principled model of protein interactions, grounded in real properties of protein family data. We introduce a simplified attention layer, factored attention, and show that it achieves comparable performance to Potts models, while sharing parameters both within and across families. Further, we extract contacts from the attention maps of a pretrained Transformer and show they perform competitively with the other two approaches. This provides evidence that large-scale pretraining can learn meaningful protein features when presented with unlabeled and unaligned data. We contrast factored attention with the Transformer to indicate that the Transformer leverages hierarchical signal in protein family databases not captured by our single-layer models. This raises the exciting possibility for the development of powerful structured models of protein family databases.1

bioinformatics↗

bioRxiv2020

Designed proteins assemble antibodies into modular nanocages

Antibodies are widely used in biology and medicine, and there has been considerable interest in multivalent antibody formats to increase binding avidity and enhance signaling pathway agonism. However, there are currently no general approaches for forming precisely oriented antibody assemblies with controlled valency. We describe the computational design of two-component nanocages that overcome this limitation by uniting form and function. One structural component is any antibody or Fc fusion and the second is a designed Fc-binding homo-oligomer that drives nanocage assembly. Structures of 8 antibody nanocages determined by electron microscopy spanning dihedral, tetrahedral, octahedral, and icosahedral architectures with 2, 6, 12, and 30 antibodies per nanocage match the corresponding computational models. Antibody nanocages targeting cell-surface receptors enhance signaling compared to free antibodies or Fc-fusions in DR5-mediated apoptosis, Tie2-mediated angiogenesis, CD40 activation, and T cell proliferation; nanocage assembly also increases SARS-CoV-2 pseudovirus neutralization by -SARS-CoV-2 monoclonal antibodies and Fc-ACE2 fusion proteins. We anticipate that the ability to assemble arbitrary antibodies without need for covalent modification into highly ordered assemblies with different geometries and valencies will have broad impact in biology and medicine.

biochemistry↗

bioRxiv2020

Design of proteins presenting discontinuous functional sites using deep learning

An outstanding challenge in protein design is the design of binders against therapeutically relevant target proteins via scaffolding the discontinuous binding interfaces present in their often large and complex binding partners. There is currently no method for sampling through the almost unlimited number of possible protein structures for those capable of scaffolding a specified discontinuous functional site; instead, current approaches make the sampling problem tractable by restricting search to structures composed of pre-defined secondary structural elements. Such restriction of search has the disadvantage that considerable trial and error can be required to identify architectures capable of scaffolding an arbitrary discontinuous functional site, and only a tiny fraction of possible architectures can be explored. Here we build on recent advances in de novo protein design by deep network hallucination to develop a solution to this problem which eliminates the need to pre-specify the structure of the scaffolding in any way. We use the trRosetta residual neural network, which maps input sequences to predicted inter-residue distances and orientations, to compute a loss function which simultaneously rewards recapitulation of a desired structural motif and the ideality of the surrounding scaffold, and generate diverse structures harboring the desired binding interface by optimizing this loss function by gradient descent. We illustrate the power and versatility of the method by scaffolding binding sites from proteins involved in key signaling pathways with a wide range of secondary structure compositions and geometries. The method should be broadly useful for designing small stable proteins containing complex functional sites.

biochemistry↗

bioRxiv2020

Computer Designed PRC2 Inhibitor, EBdCas9, Reveals Functional TATA boxes in Distal Promoter Regions

The critical process in development, bifurcation of cellular fates, requires epigenetic H3K27me3 marks propagated by PRC2 complex. However, the precise chromatin loci of functional H3K27me3 marks are not yet known. Here we identify critical PRC2 functional sites at a single nucleosome resolution. We fused a computationally designed protein, EED binder (EB) that competes with EZH2 and thereby disrupts PRC2 function, to dCas9 (EBdCas9) to direct PRC2 inhibition at a precise locus using gRNA. We targeted EBdCas9 to 4 different genes (TBX18, p16, CDX2 and GATA3) and observed epigenetic remodeling at a single nucleosome resolution resulting in gene activation. Remarkably, while traditional TATA box is located 30bp upstream of TSS, we identified a functional TATA box, >500bp of TSS, normally repressed by PRC2 complex. Deletion of this TATA box eliminates EBdCas9 dependent TBP recruitment and transcriptional activation. Targeting EBdCas9 to CDX2 and GATA3 results in trophoblast trans-differentiation. EBdCas9 technology is broadly applicable for epigenetic regulation at a single locus to control gene expression. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=189 SRC="FIGDIR/small/385922v1_ufig1.gif" ALT="Figure 1"> View larger version (62K): org.highwire.dtl.DTLVardef@1da108borg.highwire.dtl.DTLVardef@17447ddorg.highwire.dtl.DTLVardef@11a188forg.highwire.dtl.DTLVardef@426022_HPS_FORMAT_FIGEXP M_FIG C_FIG

molecular biology↗

bioRxiv2020

De novo design of transmembrane beta-barrels

The ability of naturally occurring transmembrane {beta}-barrel proteins (TMBs) to spontaneously insert into lipid bilayers and form stable transmembrane pores is a remarkable feat of protein evolution and has been exploited in biotechnology for applications ranging from single molecule DNA and protein sequencing to biomimetic filtration membranes. Because it has not been possible to design TMBs from first principles, these efforts have relied on re-engineering of naturally occurring TMBs that generally have a biological function very different from that desired. Here we leverage the power of de novo computational design coupled with a "hypothesis, design and test" approach to determine principles underlying TMB structure and folding, and find that, unlike almost all other classes of protein, locally destabilizing sequences in both the {beta}-turns and {beta}-strands facilitate TMB expression and global folding by modulating the kinetics of folding and the competition between soluble misfolding and proper folding into the lipid bilayer. We use these principles to design new eight stranded TMBs with sequences unrelated to any known TMB and show that they insert and fold into detergent micelles and synthetic lipid membranes. The designed proteins fold more rapidly and reversibly in lipid membranes than the TMB domain of the model native protein OmpA, and high resolution NMR and X-ray crystal structures of one of the designs are very close to the computational model. The ability to design TMBs from first principles opens the door to custom design of TMBs for biotechnology and demonstrates the value of de novo design to investigate basic protein folding problems that are otherwise hidden by evolutionary history. One sentence summarySuccess in de novo design of transmembrane {beta}-barrels reveals geometric and sequence constraints on the fold and paves the way to design of custom pores for sequencing and other single-molecule analytical applications.

synthetic biology↗

bioRxiv2020

Epistasis on the stability landscape of de novo TIM barrels explored by a modular design approach

The ability to design stable proteins with custom-made functions is a major goal in biochemistry with practical relevance for our environment and society. High conformational stability lowers protein sensitivity to mutations and changes in the environment; thus, understanding and manipulating protein stability will expand the applications of de novo proteins. Since the ({beta}/)8-barrel or TIM-barrel fold is one of the most common functional scaffolds, in this work we designed a collection of stable de novo TIM barrels (DeNovoTIMs), using a computational fixed-backbone and modular approach based on improved hydrophobic packing of sTIM11, the first validated de novo TIM barrel. DeNovoTIMs navigate a region of the stability landscape previously uncharted by natural TIM barrels, with variations spanning 60 degrees in melting temperature and 22 kcal per mol in conformational stability throughout the designs. Significant non-additive or epistatic effects were observed when stabilizing mutations from different regions of the barrel were combined. The molecular basis of epistasis in DeNovoTIMs appears to be related to the extension of the hydrophobic cores. This study is an important step towards the fine-tuned modulation of protein stability by design. Significance StatementDe novo protein design expands our knowledge about protein structure and stability. The TIM barrel is a highly relevant fold used in nature to host a rich variety of catalytic functions. Here, we follow a modular approach to design and characterize a collection of de novo TIM barrels and subjected them to a thorough folding analysis. Non-additive effects modulate the increase in stability when different regions of the barrel are mutated, showing a wide variety of thermodynamic properties that allow them to navigate an unexplored region of the stability landscape found in natural TIM barrels. The design of stable proteins increases the applicability of de novo proteins and provides crucial information on the molecular determinants that modulate structure and stability. One Sentence SummaryA family of designed TIM barrels with diverse thermodynamic properties shows epistatic effects on its stability landscape.

biochemistry↗

bioRxiv2020

F-domain valency determines outcome of signaling through the angiopoietin pathway

Angiopoietin 1 and 2 (Ang1 and Ang2) modulate angiogenesis and vascular homeostasis through engagement of their very similar F-domain modules with the Tie2 receptor tyrosine kinase on endothelial cells. Despite this similarity in the underlying receptor binding interaction, the two angiopoietins have opposite effects: Ang1 induces phosphorylation of protein kinase B (AKT), strengthens cell-cell junctions and enhances endothelial cell survival while Ang2 antagonizes these effects1-4. To investigate the molecular basis for the opposing effects, we examined the protein kinase activation and morphological phenotypes produced by a series of computationally designed protein scaffolds presenting the Ang1 F-domain in a wide range of valencies and geometries. We find two broad phenotypic classes distinguished by the number of presented F-domains: scaffolds presenting 4 F-domains have Ang2 like activity, upregulating pFAK and pERK but not pAKT, and failing to induce cell migration and tube formation, while scaffolds presenting 6 or more F-domains have Ang1 like activity, upregulating pAKT and inducing migration and tube formation. The scaffolds with 8 or more F-domains display superagonist activity, producing stronger phenotypes at lower concentrations than Ang1. When examined in vivo, superagonist icosahedral self-assembling nanoparticles caused significant revascularization in hemorrhagic brains after a controlled cortical impact injury.

biochemistry↗

bioRxiv2020

Design of Biologically Active Binary Protein 2D Materials

Proteins that assemble into ordered two-dimensional arrays such as S-layers1,2 and designed analogues3-5 have intrigued bioengineers,6,7 but with the exception of a single lattice formed through non-rigid template streptavidin linkers,8 they are constituted from just one protein component. For modulating assembly dynamics and incorporating more complex functionality, materials composed of two components would have considerable advantages.9-12 Here we describe a computational method to generate de-novo binary 2D non-covalent co-assemblies by designing rigid asymmetric interfaces between two distinct protein dihedral building-blocks. The designed array components are soluble at mM concentrations, but when combined at nM concentrations, rapidly assemble into nearly-crystalline micrometer-scale p6m arrays nearly identical to the computational design model in vitro and in cells without the need of a two-dimensional support. Because the material is designed from the ground up, the components can be readily functionalized, and their symmetry reconfigured, enabling formation of ligand arrays with distinguishable surfaces to drive extensive receptor clustering, downstream protein recruitment, and signaling. Using quantitative microscopy we show that arrays assembled on living cells have component stoichiometry and likely structure similar to arrays formed in vitro, suggesting that our material can impose order onto fundamentally disordered substrates like cell membranes. We find further that in sharp contrast to previously characterized cell surface receptor binding assemblies such as antibodies and nanocages, which are rapidly endocytosed, large arrays assembled at the cell surface suppress endocytosis in a tunable manner, with potential therapeutic relevance for extending receptor engagement and immune evasion. Our work paves the way towards synthetic cell biology, where a new generation of multi-protein macroscale materials is designed to modulate cell responses and reshape synthetic and living systems. One Sentence SummaryCo-assembling binary 2D protein crystals enables robust formation of complex large scale ordered biologically active materials

synthetic biology↗