Biology subjects

Beltrao, P.

Publications and source records attributed to Beltrao, P..

At least 19 recordsLinked to original sources

bioRxiv2026

A nucleolar assembly module integrates an ancestral isoaspartylase to safeguard ribosome biogenesis

Eukaryotes inherited the core ribosome biogenesis apparatus from archaea. However, nucleocytoplasmic compartmentalisation and expansion to >200 assembly factors created the challenge of integrating this ancestral machinery into a complex maturation programme. One solution is the formation of transient modules in which newly acquired assembly factors support deeply conserved components. Here, we identify such a module, centred on the ancestral isoaspartylase Fap7, which couples the modification of the ribosomal protein uS11 to its incorporation into pre-ribosomes. Fap7 partners with Krr1 to capture uS11, forming an early checkpoint in which uS11 loading licenses Kri1 engagement and assembly progression. Loss of uS11 modification triggers a late checkpoint, preventing aberrant pre-ribosomes from acquiring translational competence. Integrative structure-function studies reveal how the ancestral isoaspartylase is embedded within a conserved eukaryotic assembly-factor network to safeguard the timing, order, and fidelity of ribosome production.

biochemistry↗

bioRxiv2026

Leveraging multiplicity in biologically informed neural networks to uncover disease heterogeneity

Biologically inspired neural networks (BINNs) embed pathway, ontology, or protein-interaction structure directly into neural networks, promising interpretable disease prediction where hidden nodes map to named biological entities. Yet BINNs have been hard to train at biobank scale, and the reliability of their interpretations remains largely untested. Here we present a fast BINN implementation trained on UK Biobank genotype and plasma proteomics data from about 500,000 individuals across six common diseases. BINNs achieve competitive predictive performance, but we uncover two major limits to their interpretability. First, attribution scores are strongly biased by graph topology, because node degree and layer position influence the scores. Normalization reduces this bias but can weaken enrichment for known disease genes. Second, BINNs show substantial predictive multiplicity, that is, independently trained models with identical architecture and data reach similarly accurate solutions while prioritizing different genes and pathways. Although this multiplicity makes single-model explanations unstable, the range of interpretations can itself reveal disease biology. Across 100 replicate BINNs for type 2 diabetes, we find distinct solution clusters prioritizing either inflammatory or hepatic-metabolic pathways, mirroring known disease heterogeneity. Thus, analyzing the space of BINN explanations can turn multiplicity into a tool for studying complex disease mechanisms.

bioinformatics↗

bioRxiv2026

A Multi-Omics Framework Reveals Phosphorylation-Dependent Control of Protein Interactions

Protein phosphorylation regulates nearly every cellular process, yet most of the hundreds of thousands of human phosphosites remain functionally uncharacterized. Rather than prioritising phosphosites by conservation or structural features, here we use the abundance of an interaction partner as a readout of whether a phosphosite affects that interaction. This idea exploits the fact that subunits of stable complexes are often degraded when unbound. Here, we apply a nested linear regression model to pan-cancer data from 1,006 tumours, while controlling for transcriptional and other covariates. We identified 6,160 associations between 3,038 phosphosites and the abundance of interacting proteins, including several known interaction-regulating sites. Mapping these onto AlphaFold-predicted complexes placed 239 sites at interaction interfaces, while another 402 were linked to compartment-specific localisation, indicating that phosphorylation can also tune interactions by relocating proteins between compartments. Affinity-purification mass spectrometry of NKAP and NUF2 phosphosite mutants experimentally supported some of these predictions. Together, this framework reveals a widespread coupling between phosphorylation and interaction-dependent protein abundance and provides a prioritized, structure-informed resource for characterizing the human phosphoproteome

bioinformatics↗

bioRxiv2026

Capabilities, specificity gaps and training-data dependence of AlphaFold3 across diverse application areas

Structure prediction models have moved from single proteins to assemblies that include diverse biomolecules and their modifications. AlphaFold3 (AF3) and related models extended structural modelling via an all-atom framework, opening many new potential applications in structural biology. We evaluate how well the new capabilities of AF3 translate into application tasks in diverse areas: prediction of ubiquitinated protein structures, T-cell receptor (TCR)-epitope recognition, antibody-antigen complexes, protein-RNA and protein-lipid interactions. We find that, while AF3 can perform well in favourable settings, this performance is uneven across applications. In RNA-target predictions, the model confidence fails to separate genuine from decoy interaction partners and in several tasks accuracy depends on the presence of related complexes in the training set. Taken together, our assessment is more cautious than for AF2, whose gains in modelling monomers and complexes were clear and broadly generalisable. AF3s extension to new biomolecule types shows less consistent performance and generalisation. AF3 can be a powerful tool for hypothesis generation and prioritisation, but its predictions and use of confidence metrics will depend strongly on the specific application area and must be interpreted with respect to training-set overlap. We expect that the benchmarks provided here will serve for testing of future developments in the structure prediction field.

bioinformatics↗

bioRxiv2026

Structure of the mammalian ciliary transition zone microtubule doublet

The ciliary transition zone gates bidirectional protein trafficking to maintain the specialized ciliary proteome using microtubule doublets as a scaffold. While ciliary axonemal doublets are well-characterized, the molecular architecture of the transition zone doublet remains elusive. Here, we report the structure of the mammalian transition zone doublet from bovine tracheal cilia using cryo-electron tomography at 4.7-5.0 [A] resolution. The transition zone doublet is a structurally independent segment defined by an 8 nm-periodic arrangement of unique microtubule-inner proteins (MIPs) and microtubule-associated proteins (MAPs). We identify the calcium-binding protein CAPSL as a lumenal MIP that forms a pseudo-helical spiral and stabilizes microtubules in vitro. Furthermore, a dense MAP network on the A-microtubule surface clashes with intraflagellar transport (IFT) motor binding sites, suggesting anterograde IFT is directed to the B-microtubule for potential regulation by MAP9. Our work provides a structural framework for understanding gated ciliary transport and transition zone-linked human ciliopathies.

Cell Biology↗

bioRxiv2026

Predicting cellular functions of human phosphosites

Protein phosphorylation is a regulatory switch that controls essentially all cellular processes, including metabolism, cell cycle progression, and DNA repair. Furthermore, by modulating specific regions and activities of a protein, phosphorylation sites can independently coordinate the proteins involvement in distinct cellular processes. Despite the importance of this regulatory mechanism, the majority of human phosphosites have not been functionally characterized. Here we have developed biologically informed predictors of phosphosite cellular function by integrating protein functional annotation, recognition by upstream kinases, and regulation across phosphoproteomics perturbations. To enable this we harmonized public phosphoproteomics data into an atlas of nearly 2,000 perturbations and more than 16 million quantifications, providing a powerful resource for studying phosphosite regulation. We applied our predictors to functionally characterise thousands of understudied phosphosites, revealing many phosphosite functions that could not be inferred from protein functional annotation, and functionally divergent phosphosites within the same protein. Using cancer multi-omics data we characterized connections between these phosphosites and other molecular layers. Finally, we illuminated novel signaling axes within the DNA repair pathway, and demonstrated that phosphosites on PPM1G and KHSRP modulate cell survival under DNA damage. Our work represents a powerful resource for dissecting the cellular consequences of protein phosphorylation.

genomics↗

bioRxiv2026

Unifying the genetic landscape of common and rare diseases via latent neighborhoods

Resolving disease mechanisms and identifying safer drug targets remains challenging due to difficulties in integrating inconsistent sources of genetic evidence. Here, we developed a deep-learning (DL) method for obtaining latent representations of human diseases and other terms by coupling a variational autoencoder (VAE) to network embeddings from graph representation learning on a protein interaction network. We apply this approach to map the landscape of human disease, by systematically integrating evidence for [~]42,000 traits and terms from sources such as clinical reports (ClinVar), genome-wide association studies (GWAS), mouse phenotypes and gene ontology. Similar diseases share their latent neighborhoods with the sources of evidence highlighting different aspects of the same cell biology for its role in disease. By decoding latent neighborhoods, we unify the sources of evidence and integrate across diseases to prioritize genes for common and rare diseases. Through examples for diabetes, hearing loss, and familial long QT syndrome, we illustrate how the methodology can be applied to develop hypotheses for disease mechanisms and to propose experimental models, measurements, targets and potential drugs that may improve diagnosis and treatments.

genetics↗

bioRxiv2025

Systematic structural and functional analysis of metal-binding sites in native proteomes

Metal-binding proteins (MBPs) are widespread and critical for many biological processes. Proteomics-based approaches have identified MBPs in various systems but typically cannot determine the location of metal-binding sites or shed light on the functional roles of metals. Thus, a systems-wide map of metal binding sites and an understanding of the consequences of metal binding for protein function are still lacking. Here, we used limited proteolysis-coupled mass spectrometry (LiP-MS) combined with chelator treatment to globally map metal-protein interactions and binding sites in native protein lysates. We applied the pipeline, named Metal-LiP, to Escherichia coli, identified 981 peptides on 297 proteins that exhibited dose responsiveness to chelator treatment, and quantified the relative metal-binding affinity. These proteins are strongly enriched for known or predicted MBPs, based on our novel compiled ground truth dataset of previously annotated (n=1013) and predicted (n=850) MBPs. Further, chelator-responsive peptides mapped at or near annotated metal-binding sites in 3D protein structures, showing that Metal-LiP can identify metal-binding sites, and provided novel experimental evidence for previously predicted sites. Notably, Metal-LiP identified 225 novel candidate MBPs and we validated five of these novel MBPs (GatZ, AhpC, GrxB, MsrC, TrhO) using inductively coupled plasma-mass spectrometry to detect metal binding in purified proteins. Finally, we used orthogonal approaches to probe the broad functional consequences of metal binding. Using a combination of Metal-LiP, thermal proteome profiling, size-exclusion chromatography coupled to mass spectrometry, and structural analyses, we identified cases where metal binding affects protein stability, modulates the assembly states of protein complexes, or is likely to affect enzymatic activity. We thereby propose functional roles for metal binding in 98 of our Metal-LiP hits and in 225 proteins in total. In summary, we describe an approach to identify bona fide MBPs with binding site-level resolution across the proteome. We have combined it with other methods to identify 424 candidate MBPs in the E coli proteome, to map known and novel candidate metal-binding sites for 297 of these proteins, and to suggest functional roles of metal binding.

systems biology↗

bioRxiv2025

The functional landscape of the human ubiquitinome

Protein ubiquitination regulates cell biology through diverse avenues, from quality control-linked protein degradation to signaling functions such as modulating protein-protein interactions and enzyme activation. To date, hundreds of thousands of ubiquitination sites (ubi-sites) have been identified, however fewer than 1% have known functional roles. Here, we assembled a human reference ubiquitinome of 108,341 ubi-sites by harmonizing public proteomics data. To pinpoint critical regulatory events requiring ubiquitination at a precise site, we mapped ubi-site conservation across proteomics data from six non-human species. Perturbation proteomics revealed that highly conserved ubi-sites are more likely to regulate signaling functions rather than proteasomal degradation. To further prioritize site-specific ubiquitination relevant for organismal fitness, we constructed a machine learning-based positional importance score for more than 100,000 ubi-sites, which identifies sites regulating diverse protein functions and rationalizes genetic vulnerabilities. Finally, we employed chemical genomics to validate the functional relevance of high-scoring ubi-sites and leveraged genetic code expansion to demonstrate that ubiquitination of K320 in the RNA-regulator ELAVL1 disrupts RNA binding. Our work reveals systems-level principles of the ubiquitinome and provides a powerful resource for studying site-specific protein ubiquitination.

cell biology↗

bioRxiv2025

Predicting the protein interaction landscape of a free-living bacterium with pooled-AlphaFold3

Accurate prediction of protein complex structures by AlphaFold3 and similar programs has been used to predict the presence of protein-protein interactions (PPIs), but this technique has never been applied to an entire genome due to onerous computational requirements and questionable utility. Here we present pooled-PPI prediction, a technique that dramatically improves the accuracy of genome-scale screens compared to a paired approach while simultaneously reducing inference time ([~]2-fold) and the number of jobs ([~]100-fold). We use this technique to predict the structure of all 113,050 pairwise PPIs in Mycoplasma genitalium using only 2,027 AlphaFold3 jobs. This unbiased and comprehensive dataset was highly predictive of known interactions, revealed a previously unappreciated but widespread size bias in AlphaFold interface scores, correctly identified protein-protein interfaces in macromolecular complexes, and uncovered new biology in M. genitalium. This work establishes pooled-PPI prediction as a highly scalable method for uncovering protein-protein interactions and a powerful addition to the functional genomics toolkit.

systems biology↗

bioRxiv2025

AlphaFold models of host-pathogen interactions elucidate the prevalence and structural modes of molecular mimicry

Pathogens exploit host cellular machinery through protein-protein interactions (PPIs), often using molecular mimicry to hijack host cellular processes. While there have been thousands of host-pathogen PPIs determined to date, the lack of structural information for these impedes the study of the prevalence of molecular mimicry and convergent evolution of protein interaction interfaces. To address this, we benchmarked AlphaFold2 and 3 for prediction of structures of host-pathogen interactions observing that accurate models can be retrieved when ranking by modelling confidence, despite an overall low performance. We predicted structures for 6,782 pathogen-human PPIs yielding 803 models of higher confidence. Most pathogen proteins interacting with a common human protein are predicted to do so via the same interface, suggesting a high degree of convergent evolution of protein interaction interfaces. When comparing structural models from host-pathogen and host-host interactions, we observe that a majority of pathogen proteins are predicted to target existing human PPI interfaces. We categorized instances of mimicry into different modes, occurring at different frequencies: 1) via the same domain family (least common); 2) via a similar structural motif; and 3) via a similar linear motif (most common). We selected examples of linear motif interactions for binding assay testing, confirming 8 out of 12 predicted interfaces, including 3 viral linear motif interactions. This validates AlphaFolds ability to model some host-pathogen interactions and the mechanisms underlying molecular mimicry. This work showcases the value of large-scale structural modelling to study convergent evolution of host-pathogen interactions and how molecular mimicry may contribute to infection or host defense.

bioinformatics↗

bioRxiv2025

Network-based framework for studying etiology and phenotype diversity in primary ciliopathies

Recent advances in sequencing technologies have increasingly enabled the identification of genetic causes for human monogenic diseases. However, systematic understanding remains limited due to the rarity, genetic heterogeneity, and complex genotype-phenotype relationships of these diseases. Primary ciliopathies are a diverse group of rare disorders caused by variants in genes associated with the cilium, a cellular organelle involved in signaling during development and cell homeostasis. These genetic variants result in a wide spectrum of clinical phenotypes involving the brain, eye, kidney and skeleton. It remains unclear to what extent this phenotypic diversity can be attributed to the disease-causing genes and their specific roles in ciliary function. Here, we systematically compared human primary ciliopathies with each other and with mouse phenotypes by propagating known disease genes through a network of protein interactions. Network propagation improved the clustering of primary ciliopathies with shared clinical phenotypes and facilitated the identification of mouse phenotypes closely related to primary ciliopathies due to shared groups of proteins in the interaction network. By leveraging this phenotype-specific approach, we prioritized candidate genes for ciliopathies and identified likely pathogenic variants in CEP43, a novel candidate gene for human primary ciliopathies in three previously unsolved cases. This study demonstrates that network propagation enhances the genetic and phenotypic understanding of primary ciliopathies, aiding in the prioritization of candidate genes and identification of relevant mouse models for these rare disorders, and providing a framework for unraveling shared underlying mechanisms for other rare genetic diseases.

genetics↗

bioRxiv2024

Global comparative structural analysis of responses to protein phosphorylation

Post-translational modifications (PTMs), particularly protein phosphorylation, are key regulators of cellular processes, impacting numerous aspects of protein activity. Despite widespread phosphorylation of eukaryotic proteomes, the function of most phosphosites remains unknown. Elucidating the structural mechanisms underlying phosphorylation is crucial for understanding its regulatory roles. Here, we conducted a comparative structural analysis of phosphorylated and non-phosphorylated proteins taken from the Protein Data Bank (PDB). Our study systematically evaluates how phosphorylation affects backbone conformation, protein dynamics, and mechanical strain. We found that phosphorylation commonly induces small, stabilizing conformational changes through conformational selection and frequently modulates local residue fluctuations, influencing overall protein motion. Notably, a small but significant subset of phosphosites shows mechanical coupling with functional sites, aligning with the domino model of allosteric signal transduction. This work provides a foundation for studying phosphorylation and other PTMs in their structural context, which will guide the rational design of synthetic phosphosites and enable the engineering of PTM-driven regulatory circuits in synthetic biology.

bioinformatics↗

bioRxiv2024

Predicting the natural yeast phenotypic landscape with machine learning

Most organisms traits result from the complex interplay of many genetic and environmental factors, making their prediction from genotypes difficult. Here, we used machine learning models to explore genotype-phenotype connections for 223 life history traits measured across 1011 genome-sequenced Saccharomyces cerevisiae strains. Firstly, we used genome-wide association studies to connect genetic variants with the phenotypes. Next, we benchmarked an automated machine learning pipeline that includes preprocessing, feature selection, and hyperparameters optimization in combination with multiple linear and complex machine learning methods. We determined gradient boosting machines as best performing in 65% of predictions and pangenome as best predictor, suggesting a considerable contribution of the accessory genome in controlling phenotypes. The accuracy broadly varied among the phenotypes (r = 0.2-0.9), consistent with varying levels of complexity, with stress resistance being easier to predict compared to growth across carbon and nitrogen nutrients. While no specific genomic features could be linked to the predictions for most phenotypes, machine learning identifies high-impact variants with established relationships to phenotypes despite being rare in the population. Near-perfect accuracies (r>0.95) were achieved when other phenomics data were used to aid predictions, suggesting shared useful information can be conveyed across phenotypes. Overall, our study underscores the power of machine learning to interpret the functional outcome of genetic variants.

genetics↗

bioRxiv2024

Site-specific metabolic labeling reveals proteome-wide phospho-dynamics

Protein phosphorylation is a critical post-translational modification that orchestrates cellular signaling. Here, we introduce PulsPhos, a method combining metabolic labeling with phosphoproteomics, spectral analysis and modeling, to quantify site-specific phosphorylation lifetimes in living cells. Phosphosite lifetimes vary over multiple orders of magnitude and are influenced by factors such as amino acid composition and subcellular localization. PulsPhos was readily applied to pharmacological perturbations revealing fundamental mechanisms governing protein phosphorylation dynamics.

molecular biology↗

bioRxiv2024

Predicted mechanistic impacts of human protein missense variants

Genome sequencing efforts have led to the discovery of tens of millions of protein missense variants found in the human population with the majority of these having no annotated role and some likely contributing to trait variation and disease. Sequence-based artificial intelligence approaches have become highly accurate at predicting variants that are detrimental to the function of proteins but they do not inform on mechanisms of disruption. Here we combined sequence and structure-based methods to perform proteome-wide prediction of deleterious variants with information on their impact on protein stability, protein-protein interactions and small-molecule binding pockets. AlphaFold2 structures were used to predict approximately 100,000 small-molecule binding pockets and stability changes for over 200 million variants. To inform on protein-protein interfaces we used AlphaFold2 to predict structures for nearly 500,000 protein complexes. We illustrate the value of mechanism-aware variant effect predictions to study the relation between protein stability and abundance and the structural properties of interfaces underlying trans protein quantitative trait loci (pQTLs). We characterised the distribution of mechanistic impacts of protein variants found in patients and experimentally studied example disease linked variants in FGFR1.

bioinformatics↗

bioRxiv2024

An atlas of protein-protein associations of human tissues prioritizes candidate disease genes

Proteins that interact together participate in the same cellular process and influence the same organismal traits. Despite the progress in mapping protein-protein interactions we lack knowledge of how they differ between tissues. Due to coordinated (post)transcriptional control, protein complex members have highly correlated abundances that are predictive of functional association. Here, we have compiled 7873 proteomic samples measuring protein levels in 11 human tissues and use these to define an atlas with tissue-specific protein associations. This method recapitulates known protein complexes and the larger structural organization of the cell. Interactions of stable protein complexes are well preserved across tissues, while signaling and metabolic interactions show larger variation. Further, we find that less than 18% of differences between tissues are estimated to be due to differences in gene expression while cell-type specific cellular structures, such as synaptic components, represent a significant driver of differences between tissues. We further supported the brain protein association network through co-fractionation experiments in synaptosomes, curation of brain derived pull-down data and AlphaFold2 models. Together these results illustrate how this brain specific protein interaction network can functionally prioritize candidate genes within loci linked to brain disorders.

systems biology↗

bioRxiv2023

A foundational atlas of autism protein interactions reveals molecular convergence

Translating high-confidence (hc) autism spectrum disorder (ASD) genes into viable treatment targets remains elusive. We constructed a foundational protein-protein interaction (PPI) network in HEK293T cells involving 100 hcASD risk genes, revealing over 1,800 PPIs (87% novel). Interactors, expressed in the human brain and enriched for ASD but not schizophrenia genetic risk, converged on protein complexes involved in neurogenesis, tubulin biology, transcriptional regulation, and chromatin modification. A PPI map of 54 patient-derived missense variants identified differential physical interactions, and we leveraged AlphaFold-Multimer predictions to prioritize direct PPIs and specific variants for interrogation in Xenopus tropicalis and human forebrain organoids. A mutation in the transcription factor FOXP1 led to reconfiguration of DNA binding sites and altered development of deep cortical layer neurons in forebrain organoids. This work offers new insights into molecular mechanisms underlying ASD and describes a powerful platform to develop and test therapeutic strategies for many genetically-defined conditions.

systems biology↗