Biology subjects

Pepe, G.

Publications and source records attributed to Pepe, G..

7 recordsLinked to original sources

bioRxiv2026

LAYERED GENOMIC COMPOSITION AND LINGUISTIC CONTINUITY IN GRECIA SALENTINA

Grecia Salentina (GS) is the label assigned to a geographic area in Southern Italy where a Greek dialect has historically been spoken. The origins of this Greek-speaking community, whether dating from ancient or Medieval settlement, are a controversial and essentially unsolved issue. To evaluate which historical-linguistic scenario is best supported on genomic grounds, we used whole genome sequencing of 27 GS natives and 21 controls, embedding them in a 799 modern samples dataset in which Greece and Anatolia were evenly represented, an uncommon feature in the genomic literature on the area. We considered different aspects of the genetic structuring, at different spatial and temporal scales. Analyses also included 4021 ancient samples from the literature. All analyses pointed to the same conclusion: the GS genomic profile displays the same characteristic features of the rest of Southern Italy, with no definite evidence for a single source population from outside Italy. Correspondingly, we detected strong genetic connectivity among GS municipalities and between GS and the rest of Italy. Only a subset of our analyses suggested a slightly increased affinity of GS with Greece, or with the Anatolian peninsula. These results indicate that GS has been all but a genetically closed community, and that the Greek language did not act as a barrier to gene flow. By endorsing a ''transient mobility model'' we support the most recent linguistic hypothesis: the prolonged resilience of the Greek language in GS, maintained in situ and periodically rejuvenated through recurrent external inputs from the Greek and Byzantine worlds.

genetics↗

bioRxiv2026

Cross-Attention Over RNA And Protein Sequences Enables Generalizable Interaction Prediction

Computational predictions are essential to characterize the RNA-protein interaction landscape, yet a persistent gap between benchmark performance and practical utility suggests that current models have limited generalization capabilities. To address this issue, we present CORAL (Cross-attention for RNA-protein Association Learning), a deep learning framework for the prediction of RNA-protein interactions that integrates pretrained protein (ESM-2) and RNA (DNABERT2) language models through bidirectional cross-attention with Low-Rank Adaptation fine-tuning. We also introduce a benchmarking framework that rigorously addresses the problem of data redundancy between training and test sets, which greatly inflates model performances reported in the literature. To this end we adopt three partitioning strategies of increasing stringency: conventional random splits, pairwise non-redundant splits, and component-wise non-redundant splits. CORAL maintains an F1 score of 0.65 under the most stringent component-wise evaluation, compared to 0.47 for the next-best method retaining discriminative behavior. Interpretability analyses further reveal that the cross-attention mechanism captures biologically meaningful features of molecular recognition at two complementary structural scales: at atomic resolution, specific attention heads systematically attend to structurally defined contact positions, showing 26% elevated attention at interface residues across 309 experimentally resolved complexes (p < 0.001); at the domain level, protein-side attention localizes to annotated RNA-binding domains across 94 % of 462 proteins examined (median within-protein Cohens d {approx} 0.94). Together, these findings establish that current RPI prediction benchmarks substantially inflate performance estimates and demonstrate that cross-modal attention architectures yield improved generalization alongside mechanistically interpretable representations.

bioinformatics↗

bioRxiv2025

The Integrated Cellular and Molecular Landscape of Autoimmunity

We performed a large-scale immunogenomic analysis of [~]13,000 transcriptomic profiles from 10 autoimmune diseases, integrating publicly available datasets from both blood and tissue. Using meta-analysis, we identified core immune mechanisms underlying autoimmunity, including strong interferon responses, inflammation, and adaptive immune suppression, alongside disease-specific signatures. To enhance biological interpretability, we derived higher-order immune features - such as cell type proportions, cytokine levels, pathway activity, transcription factor regulation, and miRNA activity - and organized them into 15 coherent immune modules. These modules enabled systematic cross-disease comparisons, revealing shared and distinct immunopathological patterns. The inflammation module, in particular, was associated with disease severity and predicted treatment response across multiple conditions. This modular framework offers a powerful tool for understanding immune dysregulation and advancing precision medicine in autoimmune diseases. To support reproducibility and enable others to build upon this work, we developed an interactive app to explore and download the complete dataset and associated results.

bioinformatics↗

bioRxiv2023

Human lncRNAs harbor conserved modules embedded in different sequence contexts

We analyzed the structure of human long non-coding RNA (lncRNAs) genes to investigate whether the non-coding transcriptome is organized in modular domains, as is the case for protein-coding genes. To this aim, we compared all known human lncRNA exons and identified 340 pairs of exons with high sequence and/or secondary structure similarity but embedded in a dissimilar sequence context. We grouped these pairs in 106 clusters based on their reciprocal similarities. These shared modules are highly conserved between humans and the four great ape species, display evidence of purifying selection and likely arose as a result of recent segmental duplications. Our analysis contributes to the understanding of the mechanisms driving the evolution of the non-coding genome and suggests additional strategies towards deciphering the functional complexity of this class of molecules. Author summaryThe Human genome includes more than 18,000 genes coding for RNAs that are not translated into proteins, called long non-coding RNAs (lncRNA). Mounting evolutionary and experimental evidence shows that a large amount of these RNAs have a specific function, mainly as regulators of a diverse set of biological processes. Here we set out to investigate whether these genes have a modular organization similar to that of protein-coding genes. Accordingly, we compared the sequence of all the exonic regions of human lncRNAs and identified 106 clusters of non-repetitive exonic modules shared between this class of genes. These modules display evidence of purifying selection, are highly conserved between humans and the four great ape species, and may represent distinct functional units that have been shuffled among multiple lncRNA genes, in a manner similar to the exon-shuffling process that is observed in the coding genome.

evolutionary biology↗

bioRxiv2023

Unveiling the signaling network of FLT3-ITD AML improves drug sensitivity prediction

Currently, the identification of patient-specific therapies in cancer is mainly informed by personalized genomic analysis. In the setting of acute myeloid leukemia (AML), patient-drug treatment matching fails in a subset of patients harboring atypical internal tandem duplications (ITDs) in the tyrosine kinase domain of the FLT3 gene. To address this unmet medical need, here we develop a systems-based strategy that integrates multiparametric analysis of crucial signaling pathways, and patient-specific genomic and transcriptomic data with a prior-knowledge signaling network using a Boolean-based formalism. By this approach, we derive personalized predictive models describing the signaling landscape of AML FLT3-ITD positive cell lines and patients. These models enable us to derive mechanistic insight into drug resistance mechanisms and suggest novel opportunities for combinatorial treatments. Interestingly, our analysis reveals that the JNK kinase pathway plays a crucial role in the tyrosine kinase inhibitor response of FLT3-ITD cells through cell cycle regulation. Finally, our work shows that patient-specific logic models have the potential to inform precision medicine approaches.

cancer biology↗

bioRxiv2023

NRF2 connects Src tyrosine kinase to ferroptosis resistance in glioblastoma

Glioblastoma (GBM) is a severe brain tumor characterized by an extremely poor survival rate of patients. GBM cancer cells escape to standard therapeutic protocols consisting of combination of ionizing radiation (IR) and alkylating drugs that trigger DNA damage, by rewiring of signaling pathways. In recent years, the upregulation of factors that counteract ferroptosis has been highlighted as a major driver of cancer resistance to IR, although the molecular connection between the activation of oncogenic signaling and the modulation of ferroptosis has not been clarified yet. Here we provide the first evidence for a molecular connection between the constitutive activation of tyrosine kinases and resistance to ferroptosis. Src tyrosine kinase, a central hub on which Receptor Tyrosine Kinases deregulated signaling converge in cancer, leads to the stabilization and activation of NRF2 pathway, thus promoting resistance to IR-induced ferroptosis. These data suggest that the upregulation of Src-NRF2 axis may represent a vulnerability for combined strategies that, by targeting ferroptosis resistance, enhance radiation sensitivity in glioblastoma.

cancer biology↗

bioRxiv2021

Adaptive maternal effects in early life history traits help to maintain ecological resilience in novel environments for two contrasting Senecio species

While many organisms shift their development plastically to maintain fitness as environments change, such plasticity has limits. Population mean fitness is expected to decline when novel environments exceed the limits to plasticity, but the level of fitness costs is expected to vary among genotypes that can increase adaptive potential. We lack fundamental insights into how and when changes in early development traits increase adaptive potential in novel environments, which limits our ability to predict the response of natural populations to global change. To test whether genetic variation in development time is associated with increased adaptive potential in novel environments, we used a breeding design to generate c.20,000 seeds of two ecologically contrasting Sicilian species of daisies (Senecio, Asteraceae) adapted to high and low elevations on Mount Etna. We planted the seeds across four elevations that included the native range of each species, and a novel elevation. We tracked seedling mortality and measured development time as the number of days it took seedlings to establish. As predicted, genetic variance in survival increased at novel elevations. However, genetic variance in development time showed the opposite trend, decreasing at novel elevations. A strong negative genetic correlation between development time in the native range and survival at novel elevations suggested that genotypes with faster development in native environments survived better in novel environments. These results were consistent across the two ecologically contrasting species, suggesting that genetic variance in early development in native environments could be used to predict population responses to novel environments.

evolutionary biology↗