Biology subjects

Carbone, A.

Publications and source records attributed to Carbone, A..

At least 19 recordsLinked to original sources

bioRxiv2026

Residue-level predictions of the protein-protein interactions of the hepatitis B virus core and envelope proteins

We here predicted the interactions between the capsid (Cp) and envelope proteins (S/M/LHBs) of the hepatitis B virus using a recently established mutation-driven deep-learning model, as well as coevolution signatures that serve as markers of physical interactions and/or functional relationships. The sequence-based analyses reveal putative protein-protein interaction (PPI) hotspots in proteins, and identify abundant coevolved residues within and across proteins. We analyze the results with a focus on the intermolecular interactions between Cp and the large envelope protein LHBs, especially its disordered preS domain. We compare the predicted PPI interface sites to previous evidence on PPIs, derived from mutational analyses described in the literature. We equally integrate experimental NMR data that provide a rationale for the previous observation that spike-binding peptides inhibit core-envelope interactions. Our work sheds new light on the molecular mechanisms at play on HBV envelopment, and provides starting points for the experimental investigation of these interactions using structural and molecular virology approaches.

biophysics↗

bioRxiv2026

PLMView: collaborative protein language model representations for fast and scalable specialized protein function inference

The functional classification of protein sequences remains a major bottleneck in biology. Although protein language model (PLM)-based approaches have substantially improved broad protein function prediction, most protein sequences still lack precise annotation at the level of specialized functions--the fine-grained molecular roles that define specificity within protein families. We present PLMView, an unsupervised framework for fine-grained protein function classification directly from sequence. PLMView reframes protein function inference as a relational problem: instead of embedding sequences in isolation, it positions them within a collaborative functional space defined by comparisons with PLM embeddings of anchor sequences, thereby capturing subtle sequence-function relationships. Without requiring labeled data, family-specific training, or PLM fine-tuning, PLMView accurately distinguishes specialized functions among homologous proteins and highlights residues likely to determine functional specificity. The method achieves high precision while remaining computationally efficient, classifying approximately 10,000 sequences with 1,000 anchors in under 40 minutes; compared with pooled-embedding approaches and, in challenging cases, Sequence Similarity Networks, PLMView provides finer and more biologically coherent functional resolution, while achieving more than 10-fold speed-up over SSN reconstruction on datasets of this scale. Applications to thioredoxins, visual opsins, and Tara Oceans environmental diatom cold-shock proteins show that PLMView can move from interpretable residue-level determinants in well-studied protein families to large-scale environmental functional discovery, linking molecular specialization to ecological distribution and transcriptional deployment across the global ocean.

bioinformatics↗

bioRxiv2026

X-PAIR: an ultrafast multitask framework for proteome-scale reconstruction of PPI networks and partner-specific interfaces from sequence

Protein-protein interaction prediction and residue-level interface localisation are biologically intertwined but usually treated as separate computational problems. Here we present X-PAIR, a sequence-based multitask deep learning framework that jointly predicts whether two proteins interact and identifies their partner-specific interface residues. By combining protein language-model representations with lightweight cross-attention, X-PAIR requires neither structural templates nor multiple-sequence alignments. Across leakage-controlled benchmarks, it outperforms existing methods in both tasks, with substantial gains in interface localisation. Multitask learning preserves single-task performance while returning both outputs at near-single-task cost, enabling one million protein pairs to be analysed in under two hours--approximately 500-fold faster for interface prediction and 20-fold faster for PPI prediction than current approaches--thereby enabling proteome-scale analysis. Cross-species analyses reveal distinct evolutionary dependencies: interaction prediction benefits from multispecies training, whereas interface localisation remains robust across taxonomic scales. X-PAIR thus links proteome-scale interaction discovery to the residue-level determinants of partner-specific molecular recognition.

bioinformatics↗

bioRxiv2026

Inflammatory injury drives CIN and STING loss in polyploidpancreatic acinar cells

Epidemiological studies link pancreatic inflammation to increased cancer risks with elevated levels of chromosomal instability (CIN), yet the underlying cause remains unclear. The exocrine pancreas and lactating mammary gland harbor polyploid cells to boost secretory efficiency. Using organoid and in vivo models of pancreatic injury, we show that polyploid acinar cells contribute to regeneration via acinar-to-ductal metaplasia. The metaplasia induced cell size reduction is associated with crowding and spindle misorientation, leading to mitotic errors such as lagging chromosomes, chromatin bridges, and micronuclei. Micronucleated cells are not always eliminated or arrested in the cell cycle but can continue to proliferate and expand. Polyploid cells in post-lactational mammary gland organoids equally exhibited proliferation-related CIN. Single-cell proteomics of acinar cells shows STING1 loss and increased ATR levels in polyploid cells, enabling continued proliferation. These findings uncover physiological links between polyploidy, tissue repair, and CIN, offering new insights into cancer risks after tissue remodeling.

cell biology↗

bioRxiv2025

Genomic architecture of the self-incompatibility locus in apple provides insights into the evolution of collaborative non-self recognition

Self-incompatibility (SI) systems prevent self-fertilization, thereby maintaining genetic diversity in flowering plants. Among them, collaborative non-self recognition (CNSR) is the most widespread, yet the genomic organization and evolutionary maintenance of its multigenic recognition system remain poorly understood. Using 27 haplotype-resolved genomes from wild and domesticated apples (Malus spp.), we dissected the structure and evolution of the S-locus. We identified 17 S-RNase alleles and 500 pollen-expressed S-locus F-box brother (SFBB) genes across 18 families. The S-locus shows extensive structural divergence among alleles and transposable element accumulation, consistent with long-term restricted effective recombination. Despite this divergence, haplotypes carrying the same S-RNase allele retain remarkably conserved SFBB repertoires and gene organization, even across species boundaries, indicating that long-term balancing selection preserves highly conserved S-haplotype architectures associated with specific S-RNase lineages. Tandem duplication, positive selection, and signatures consistent with gene conversion contribute to the diversification of pollen-expressed SFBB genes while S-RNase-associated SFBB repertoires remain conserved across haplotypes carrying the same S-RNase allele. Our results reveal how a structurally dynamic yet evolutionarily constrained genomic region can sustain long-term S-allele diversity and preserve complex multigenic haplotype architectures in flowering plants.

genomics↗

bioRxiv2025

LoRA-DR-suite: adapted embeddings predict intrinsic and soft disorder from protein sequences

Intrinsic disorder regions (IDR) and soft disorder regions (SDR) provide crucial information on a protein structure to underpin its functioning, interaction with other molecules and assembly path. Circular dichroism experiments are used to identify intrinsic disorder residues, while SDRs are characterized using B-factors, missing residues, or a combination of both in alternative X-ray crystal structures of the same molecule. These flexible regions in proteins are particularly significant in diverse biological processes and are often implicated in pathological conditions. Accurate computational prediction of these disordered regions is thus essential for advancing protein research and understanding their functional implications. To address this challenge, LoRA-DR-suite employs a simple adapter-based architecture that utilizes protein language models embeddings as protein sequence representations, enabling the precise prediction of IDRs and SDRs directly from primary sequence data. Alongside the fast LoRA-DR-suite implementation, we release SoftDis, a unique soft disorder database constructed for approximately 500,000 PDB chains. SoftDis is designed to facilitate new research, testing, and applications on soft disorder, advancing the study of protein dynamics and interactions.

bioinformatics↗

bioRxiv2024

Comprehensive Mutational Landscape Analysis of Monkeypox Virus Proteome

In this study, we present a comprehensive computational analysis of the single point mutational landscapes of the Monkeypox virus (MPXV) proteome. We reconstructed full single-point mutational landscapes of 171 MPXV proteins using an advanced mutational effect predictor, ESCOTT, selected for its superior performance on viral proteins. ESCOTT performance was assessed by benchmarking against the experimental data in the ProteinGym (v1.0.0) dataset that contains 48917 multiple and 173502 single point mutations. A recent MPXV strain sequenced in July 2024 was used as the reference genome. Multiple sequence alignments and protein structures were generated using Colabfold v1.5.5, and the predicted structures were evaluated with pLDDT metric, secondary structure predictions, and comparisons with available experimental data, ensuring high confidence in the structural models. We determined mutational sensitivity of all positions in a protein utilizing ESCOTT scores and demonstrated their functional implications on cysteine proteinase and helicase of MPXV. Moreover, we created an interactive visualization tool to visualize mutational landscapes and sensitivities in a publicly available Google Colab. Furthermore, we introduced a novel, interpretable metric (Average Gene Mutation Sensitivity) to prioritize the most mutation-sensitive proteins within the large MPXV proteome as prime candidates for drug or vaccine development. Among the top 20 proteins identified with this metric, several were membrane-associated proteins, proven to be important for viral interactions with the hosts in other viruses. This analysis provides a valuable resource for assessing the impact of new MPXV variants. This pioneering study underscores the significance of understanding MPXV evolution in the context of the ongoing global health crisis and offers a robust computational framework to support this effort.

microbiology↗

bioRxiv2024

MuLAN: Mutation-driven Light Attention Networks for investigating protein-protein interactions from sequences

Understanding how proteins interact and how mutations affect these interactions is crucial for unraveling the complexities of biological systems and their evolution. Mutations can significantly alter protein behavior, impacting stability, interactions, and activity, thereby affecting cellular functions and influencing disease development and treatment effectiveness. Experimental methods for examining protein interactions are often slow and costly, highlighting the need for efficient computational strategies. We present MuLAN, a groundbreaking deep learning method that leverages light attention networks and the power of pre-trained protein language models to infer protein interactions, predict binding affinity changes, and reconstruct mutational landscapes for proteins involved in binary interactions, starting from mutational changes and directly using sequence data only. Unlike previous methods that depend heavily on structural information, MuLANs sequence-based approach offers faster and more accessible predictions. This innovation allows for variations in predictions based on specific partners, opening new possibilities for understanding protein behavior through their sequences. The potential implications for disease research and drug development mark a significant step forward in the computational analysis of protein interactions.

bioinformatics↗

bioRxiv2024

Unlocking Protein Evolution Insights: Efficient and Interpretable Mutational Effect Predictions with GEMME

The Global Epistatic Model for predicting Mutational Effects (GEMME) is a computational method that reconstructs protein mutational landscapes using sequence data alone. In this article, we delve into the broader biological questions that GEMME can address beyond mere landscape reconstruction. We provide several examples to guide users in maximizing the utility of GEMME. Additionally, we discuss recent advancements that enhance GEMMEs predictive accuracy and extend its capabilities by integrating sequence data with structural information and allele frequency data, when available. These innovations enable GEMME to answer new biological questions and offer deeper insights into protein evolution and function.

bioinformatics↗

bioRxiv2023

SENSE-PPI reconstructs protein-protein interactions of various complexities, within, across, and between species, with sequence-based evolutionary scale modeling and deep learning

Ab initio computational reconstructions of protein-protein interaction (PPI) networks will provide invaluable insights on cellular systems, enabling the discovery of novel molecular interactions and elucidating biological mechanisms within and between organisms. Leveraging latest generation protein language models and recurrent neural networks, we present SENSE-PPI, a sequence-based deep learning model that efficiently reconstructs ab initio PPIs, distinguishing partners among tens of thousands of proteins and identifying specific interactions within functionally similar proteins. SENSE-PPI demonstrates high accuracy, limited training requirements, and versatility in cross-species predictions, even with non-model organisms and human-virus interactions. Its performance decreases for phylogenetically more distant model and non-model organisms, but signal alteration is very slow. SENSE-PPI is state-of-the-art, outperforming all existing methods. In this regard, it demonstrates the important role of parameters in protein language models. SENSE-PPI is very fast and can test 10,000 proteins against themselves in a matter of hours, enabling the reconstruction of genome-wide proteomes. Graphical abstractSENSE-PPI is a general deep learning architecture predicting protein-protein interactions of different complexities, between stable proteins, between stable and intrinsically disordered proteins, within a species, and between species. Trained on one species, it accurately predicts interactions and reconstructs complete specialized subnetworks for model and non-model organisms, and trained on human-virus interactions, it predicts human-virus interactions for new viruses. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=187 SRC="FIGDIR/small/558413v3_ufig1.gif" ALT="Figure 1"> View larger version (63K): org.highwire.dtl.DTLVardef@106d345org.highwire.dtl.DTLVardef@11883a6org.highwire.dtl.DTLVardef@6b07d9org.highwire.dtl.DTLVardef@d07138_HPS_FORMAT_FIGEXP M_FIG C_FIG

bioinformatics↗

bioRxiv2023

A layer cake model for plant and metazoan chromatin

Early structural studies of chromosome organisation in eukaryotic nuclei led to the identification of euchromatin and heterochromatin, two main archetypes of chromatin visible through microscopy. Subsequently, the diversity of chromatin composition along the linear genome sequence was resolved using high throughput sequencing techniques to map a multitude of chromatin marks and unravel their functional organisation. Recent analyses of human chromosome conformation capture experiments tend to point to the existence of several chromosome sub-compartments that may correspond to epigenomic variations in chromatin composition. Here, we compare genome 3D organisations in representative eukaryotic species to explore the links between chromosomal sub-compartments and chromatin marks, genome replication timing, and genomic repeats in six model organisms, including vertebrates, plants and insects. We report that the 3D organisation of chromatin in organisms with different genome content and size can be described as layers characterised by distinct chromatin marks and activities. We propose a "layer cake" model for the genome 3D organisation as a more refined view than the prevalent "two compartments" model of chromatin organisation in multi-cellular organisms.

genomics↗

bioRxiv2022

Deep Local Analysis deconstructs protein-protein interfaces and accurately estimates binding affinity changes upon mutation

AO_SCPLOWBSTRACTC_SCPLOWThe spectacular recent advances in protein and protein complex structure prediction hold promise for reconstructing interactomes at large scale and residue resolution. Beyond determining the 3D arrangement of interacting partners, modeling approaches should be able to unravel the impact of sequence variations on the strength of the association. In this work, we report on Deep Local Analysis (DLA), a novel and efficient deep learning framework that relies on a strikingly simple deconstruction of protein interfaces into small locally oriented residue-centered cubes and on 3D convolutions recognizing patterns within cubes. Merely based on the two cubes associated with the wild-type and the mutant residues, DLA accurately estimates the binding affinity change for the associated complexes. It achieves a Pearson correlation coefficient of 0.81 on more than 2 000 mutations, and its generalization capability to unseen complexes is higher than the state-of-the-art methods. We show that taking into account the evolutionary constraints on residues contributes to predictions. We also discuss the influence of conformational variability on performance. Beyond the predictive power on the effects of mutations, DLA is a general framework for transferring the knowledge gained from the available non-redundant set of complex protein structures to various tasks. For instance, given a single partially masked cube, it recovers the identity and physico-chemical class of the central residue. Given an ensemble of cubes representing an interface, it predicts the function of the complex. Source code and models are available at http://gitlab.lcqb.upmc.fr/DLA/DLA.git.

bioinformatics↗

bioRxiv2022

Deep Local Analysis estimates effects of mutations on protein-protein interactions

The spectacular advances in protein and protein complex structure prediction hold promises for the reconstruction of interactomes at large scale at the residue resolution. Beyond determining the 3D arrangement of interacting partners, modeling approaches should be able to sense the impact of sequence variations such as point mutations on the strength of the association. In this work, we report on DLA-mutation, a novel and efficient deep learning framework for accurately predicting mutation-induced binding affinity changes. It relies on a 3D-invariant description of local 3D environments at protein interfaces and leverages the large amounts of available protein complex structures through self-supervised learning. It combines the learnt representations with evolutionary information, and a description of interface structural regions, in a siamese architecture. DLA-mutation achieves a Pearson correlation coefficient of 0.81 on a large collection of more than 2000 mutations, and its generalization capability to unseen complexes is higher than state-of-the-art methods.

bioinformatics↗

bioRxiv2022

Tumoroscope: a probabilistic model for mapping cancer clones in tumor tissues

Spatial and genomic heterogeneity of tumors is the key for cancer progression, treatment, and survival. However, a technology for direct mapping the clones in the tumor tissue based on point mutations is lacking. Here, we propose Tumoroscope, the first probabilistic model that accurately infers cancer clones and their high-resolution localization by integrating pathological images, whole exome sequencing, and spatial transcriptomics data. In contrast to previous methods, Tumoroscope explicitly addresses the problem of deconvoluting the proportions of clones in spatial transcriptomics spots. Applied to a reference prostate cancer dataset and a newly generated breast cancer dataset, Tumoroscope reveals spatial patterns of clone colocalization and mutual exclusion in sub-areas of the tumor tissue. We further infer clone-specific gene expression levels and the most highly expressed genes for each clone. In summary, Tumoroscope enables an integrated study of the spatial, genomic, and phenotypic organization of tumors.

cancer biology↗

bioRxiv2022

Deep Local Analysis evaluates protein docking conformations with locally oriented cubes

AO_SCPLOWBSTRACTC_SCPLOWWith the recent advances in protein 3D structure prediction, protein interactions are becoming more central than ever before. Here, we address the problem of determining how proteins interact with one another. More specifically, we investigate the possibility of discriminating near-native protein complex conformations from incorrect ones by exploiting local environments around interfacial residues. Deep Local Analysis (DLA)-Ranker is a deep learning framework applying 3D convolutions to a set of locally oriented cubes representing the protein interface. It explicitly considers the local geometry of the interfacial residues along with their neighboring atoms and the regions of the interface with different solvent accessibility. We assessed its performance on three docking benchmarks made of half a million acceptable and incorrect conformations. We show that DLA-Ranker successfully identifies near-native conformations from ensembles generated by molecular docking. It surpasses or competes with other deep learning-based scoring functions. We also showcase its usefulness to discover alternative interfaces. Availabilityhttp://gitlab.lcqb.upmc.fr/dla-ranker/DLA-Ranker.git

bioinformatics↗

bioRxiv2022

Soft disorder modulates the assembly path of protein complexes

The relationship between interactions, flexibility and disorder in proteins has been explored from many angles over the years: folding upon binding, flexibility of the core relative to the periphery, entropy changes, etc. In this work, we provide statistical evidence for the involvement of highly mobile and disordered regions in complex assembly. We ordered the entire set of X-ray crystallographic structures in the Protein Data Bank into hierarchies of progressive interactions involving identical or very similar protein chains, yielding 40205 hierarchies of protein complexes with increasing numbers of partners. We then examine them as proxies for the assembly pathways. Using this database, we show that upon oligomerisation, the new interfaces tend to be observed at residues that were characterised as softly disordered (flexible, amorphous or missing residues) in the complexes preceding them in the hierarchy. We also rule out the possibility that this correlation is just a surface effect by restricting the analysis to residues on the surface of the complexes. Interestingly, we find that the location of soft disordered residues in the sequence changes as the number of partners increases. Our results show that there is a general mechanism for protein assembly that involves soft disorder and modulates the way protein complexes are assembled. This work highlights the difficulty of predicting the structure of large protein complexes from sequence and emphasises the importance of linking predictors of soft disorder to the next generation of predictors of complex structure. Finally, we investigate the relationship between the Alphafold2s confidence metric pLDDT for structure prediction in unbound versus bound structures, and soft disorder. We show a strong correlation between Alphafold2 low confidence residues and the union of all regions of soft disorder observed in the hierarchy. This paves the way for using the pLDDT metric as a proxy for predicting interfaces and assembly paths. AvailabilityAll the data used for these analyses (hierarchies of interactions and soft disorder) are available at the website http://www.lcqb.upmc.fr/softdisorder-assembly/. Author summaryBoth flexibility and intrinsic disorder are used as regulatory mechanisms in proteins. They can alter the spatial positions of important recognition sites, and increased mobility appears to facilitate ligand binding through conformational selection. In this work, we show statistical evidence that soft disorder is directly involved in the process of protein assembly and that migration of soft disorder after binding gives rise to new or altered functions in the protein complex. Given the impressive progress that AlphaFold2 has made in protein structure prediction in recent years, this work highlights the importance of also correctly predicting conformational heterogeneity, mobility and intrinsic disorder in order to access the full functional repertoire and interaction network of a given protein.

bioinformatics↗

bioRxiv2021

Crystal structure of chloroplast fructose-1,6-bisphosphate aldolase from the green alga Chlamydomonas reinhardtii

The Calvin-Benson cycle fixes carbon dioxide into organic triosephosphates through the collective action of eleven conserved enzymes. Regeneration of ribulose-1,5-bisphosphate, the substrate of Rubisco-mediated carboxylation, requires two lyase reactions catalyzed by fructose-1,6-bisphosphate aldolase (FBA). While cytoplasmic FBA has been extensively studied in non-photosynthetic organisms, functional and structural details are limited for chloroplast FBA encoded by oxygenic phototrophs. Here we determined the crystal structure of plastidial FBA from the unicellular green alga Chlamydomonas reinhardtii (Cr). We confirm that CrFBA folds as a TIM barrel, describe its catalytic pocket and homo-tetrameric state. Multiple sequence profiling classified the photosynthetic paralogs of FBA in a distinct group from non-photosynthetic paralogs. We mapped the sites of thiol- and phospho-based post-translational modifications known from photosynthetic organisms and predict their effects on enzyme catalysis.

biochemistry↗

bioRxiv2021

Dietary adaptation in Neandertal, Denisovan and Sapiens revealed by gene copy number variation

Dietary adaptation involves evolving an efficient system to digest food available in an ecosystem. The diet of archaic humans is traditionally reconstructed by isotopic analyses of human remains combined with the faunal assemblages found on the sites, and, recently, from metagenomic analyses of dental calculus. Here, we propose a new computational approach to find the genetic basis for human dietary adaptation. We searched 15 genomes from Neandertal, Denisovan and Early Sapiens for food digestion genes that tend to have more or fewer copies than the modern human reference genome. We identify 50 genes, including 10 gene clusters, with discernible copy number variation (CNV) trends at the population level, from an analysis of the full set of 20,000 human genes. The genomic variation of 19 of these genes shows how metabolic pathways for carbohydrates, lipids, liver lipids and brown fat in archaic humans adapted to metabolize food from animal or plant sources. The remaining 31 genes are all highly expressed in tissues of the digestive apparatus and are involved in immune response, environmental response and obesity. Analysis of the CNV profiles, compared to 64 modern human individuals belonging to distinct ethnic groups in Eurasia, Africa, Oceania, suggests that Homo sapiens may have had an evolutionary advantage compared to Neandertal and Denisovan in adapting to cold and temperate ecosystems. Significance StatementUnderstanding dietary strategies and foraging behaviors of past populations are among the goals of paleoanthropological and prehistoric studies. Based on a new computational approach, we seek for the genetic basis for human dietary adaptation. Gene copy number variation (CNV) is a major type of structural genome variation that we used as indicator of efficient metabolic processes for food digestion. By analysing billions of short sequences from 15 archaic human genomes and 64 modern human ones, across the full set of 20,000 human genes, we identify 50 genes whose population-wide discernable CNV trends point to lipid metabolism as being crucial for Neandertal, the efficiency in carbohydrate metabolism for Sapiens diet, and the importance of lipid metabolism and brown fat metabolism for both Neandertal and early Sapiens compared to modern humans. Preprint availability: an early version, where the CNV analysis of ancient genomes was conducted on the human reference genome hg19, was deposited on BioRxiv on November 2, 2021 (DOI: 10.1101/2021.10.30.466563). The current version, which utilises CNV estimates based on the human reference genome hg38 and includes comparisons with modern human populations, was deposited in BioRxiv in November 2024.

genomics↗