Biology subjects

Porta-Pardo, E.

Publications and source records attributed to Porta-Pardo, E..

15 recordsLinked to original sources

bioRxiv2026

RNA and protein are not interchangeable in single-cell spatial multi-omics

Spatial transcriptomics is increasingly used to infer cellular phenotypes, molecular interactions, and functional tissue organization from RNA, although most of these biological processes are ultimately mediated by proteins. Here, we directly tested the interchangeability of spatial RNA and protein measurements using paired Xenium profiling of 21 matched targets across ~1.8 million cells from primary lung adenocarcinoma and metastatic castration-resistant prostate cancer. After accounting for marker- and dataset-specific protein background, we quantified concordance at multiple levels, including single-cell detection, expression magnitude, cell-type distribution, patient-level abundance, and spatial organization. Agreement at single-cell level was strongly marker-dependent. Canonical lineage markers generally showed concordant RNA and protein patterns, whereas several molecules, including in our study PD-1, PD-L1, and LAG-3, showed little concordance across modalities. Importantly, spatial organization remained substantially more discordant between modalities than overall abundance, indicating that spatial biomarkers are not directly transferable across modalities. Together, our results show that spatial RNA and protein measurements provide largely overlapping information for lineage markers and overall abundance, but can diverse for specific molecules and for higher-order spatial features, where the two layers provide complementary rather than equivalent readouts. These findings provide practical guidance on when spatial transcriptomic readouts can be interpreted as proxies for protein and when orthogonal protein validation is advisable.

cancer biology↗

bioRxiv2026

A pan-cancer single-cell atlas of pericytes

Pericytes display marked tissue-specific transcriptional identities, raising the question of whether tumour-associated pericytes converge towards shared adaptive states across cancers. Here, we built a pan-cancer single-cell RNA sequencing atlas of pericytes, integrating nearly four million cells across nine tissues, complemented by spatial transcriptomic analyses. Despite their physiological diversity, pericytes were recurrently expanded in tumours and converged on a common transcriptional program, the tumour-associated pericyte signature (TAPS). TAPS robustly identified pericytes across datasets, outperforming canonical markers in tumour contexts. Tumour-associated pericytes further diversified into specialised states including extracellular matrix (ECM)-associated and interferon (IFN)-responsive programs, which occupy mutually exclusive tumour ecosystems. ECM-associated pericytes were enriched in desmoplastic, fibroblast-rich regions and were associated with adverse clinical outcomes across multiple cancer types. IFN-responsive pericytes accumulated in inflammatory niches, with macrophages implicated as candidate drivers of specialisation. Together, our multi-layered analysis defines convergent and specialised tumour-associated pericyte programs across human cancers.

cancer biology↗

bioRxiv2026

Convergent stromal and immune remodeling defines spatial tumor dynamics in PARP inhibitor-resistant high-grade serous ovarian cancer

Acquired resistance to poly(ADP-ribose) polymerase inhibitors (PARPi) remains a major barrier to durable clinical benefit in patients with high-grade serous ovarian cancer (HGSOC), yet the profiles of tumors at resistance remain poorly understood. We profiled longitudinal HGSOC patient tumors spanning diagnosis, post-neoadjuvant therapy, and progression on PARPi with complementary spatial transcriptomics platforms, integrating single-cell-resolution Xenium analysis in paired longitudinal cases with cohort-level Visium profiling. PARPi failure was associated with spatial remodeling of the tumor microenvironment, marked by hypoxia-associated malignant transcriptional programs, increased stromal compartmentalization, and exclusion of effector immune cells. Together, these findings indicate that PARPi resistance in HGSOC is accompanied by reproducible spatial niche remodeling, underscoring the tumor microenvironment as a key contextual determinant of therapeutic failure.

cancer biology↗

bioRxiv2024

Spatial transcriptomics unveils the in situ cellular and molecular hallmarks of the lung in fatal COVID-19

Severe Coronavirus disease 2019 (COVID-19) induces heterogeneous and progressive diffuse alveolar damage (DAD) highly disrupting lung tissue architecture and homeostasis, hampering disease management leading to fatal outcomes. Characterizing DAD pathophysiology across disease progression is of ultimate importance to better understand the molecular and cellular features driving different DAD patterns and to optimize treatment strategies. To contextualize the interplay between cell types and assess their distribution, spatial transcriptomics (ST) techniques have emerged, allowing unprecedented resolution to investigate spatial architecture of tissues. To this end, post-mortem lung tissue provides valuable insights into cellular composition and their spatial relationships at the time of death. Here, we have leveraged VisumST technology in post-mortem COVID-19 induced acute and proliferative DAD lungs including control samples with normal morphological appearance, to unravel the immunopathological mechanisms underlying DAD, providing novel insights into cellular and molecular communication events driving DAD progression in fatal COVID-19. We report a progressive loss of endothelial cell types, pneumocytes type I and natural killer cells coupled with a continuous increase of myeloid and stromal cells, mostly peribronchial fibroblasts, over disease progression. Spatial organization analysis identified variable cellular compartments, ranging from major compartments defined by cell type lineages in control lungs to increased and more specific compartmentalization including immune-specific clusters across DAD spectrum. Importantly, spatially informed ligand-receptor interaction (LRI) analysis revealed an intercellular communication signature defining COVID-19 induced DAD lungs. Transcription factor (TF) activity enrichment analysis identified TGF-B pathway as DAD driver, highlighting SMAD3 and SMAD7 TFs activity role during lung fibrosis. Integration of deregulated LRIs and TFs activity allowed us to propose a downstream intracellular signaling pathway in peribronchial fibroblasts, suggesting potential novel therapeutic targets. Finally, spatio-temporal trajectories analysis provided insights into the alveolar epithelium regeneration program, characterizing markers of pneumocytes type II differentiation towards pneumocytes type I. In conclusion, we provide a spatial characterization of lung tissue architecture upon COVID-19 induced DAD progression, identifying molecular and cellular hallmarks that may help optimize treatment and patient management.

pathology↗

bioRxiv2024

Comparison of spatial transcriptomics technologies across six cancer types

Spatial biology experiments integrate the molecular and histological landscape of tissues to provide a previously inaccessible view of tissue biology, unlocking the architecture of complex multicellular tissues. Within spatial biology, spatial transcriptomics platforms are among the most advanced, allowing researchers to characterize the expression of thousands of genes across space. These new technologies are transforming our understanding of how cells are organized in space and communicate with each other to determine emergent phenotypes. This is particularly important in cancer research, as tumor evolution is shaped not only by the genetic properties of cancer cells but also by how they interact with the tumor microenvironment and their spatial organization. While many platforms can generate spatial transcriptomics profiles, it is still unclear in which context each platform better suits the needs of its users. Here we compare the results obtained using 5 different spatial transcriptomics (VISIUM, VISIUM CytAssist, VisiumHD, Xenium, and CosMx) and one spatial proteomics (VISIUM CytAssist) platforms across serial sections of 6 FFPE samples from primary human tumors covering some of the most common forms of the disease (lung, breast, colorectal, bladder, lymphoma and ovary). We observed that the VISIUM platform with CytAssist chemistry yielded superior data quality than manual VISIUM. On the other hand, Xenium consistently produced more reliable results for in situ platforms, with better gene clustering and fewer false positives than CosMx. Importantly, these platform-based variations didnt significantly affect cell type identification. VisiumHD offers the best options of both worlds, with subcellular resolution and whole-transcriptome coverage, albeit with some limitations that need to be accounted for in downstream analyses. Finally, by comparing VISIUM protein profiles with the spatial transcriptomics data from all four platforms on each sample, we identified several genes with mismatched RNA and protein expression patterns, highlighting the importance of multi-omics profiling to reveal the true biology of human tumors.

genomics↗

bioRxiv2023

3Dmapper: A Command Line Tool For BioBank-scale Mapping Of Variants To Protein Structures

The interpretation of genomic data is crucial to understand the molecular mechanisms of biological processes. Protein structures play a vital role in facilitating this interpretation by providing functional context to genetic coding variants. However, mapping genes to proteins is a tedious and error-prone task due to inconsistencies in data formats. Over the past two decades, numerous tools and databases have been developed to automatically map annotated positions and variants to protein structures. However, most of these tools are web-based and not well-suited for large-scale genomic data analysis. To address this issue, we introduce 3Dmapper, a standalone command-line tool developed in Python and R. It systematically maps annotated protein positions and variants to protein structures, providing a solution that is both efficient and reliable. 3Dmapper is freely available on GitHub at https://github.com/vicruiser/3Dmapper.

bioinformatics↗

bioRxiv2023

Effect of Aging on the Human Myometrium at Single-Cell Resolution

The myometrial dysfunction associated with aging can prompt complications during pregnancy and labor, causing a 7.8-fold increase in maternal mortality in women over 40. Using single-cell/single-nucleus RNA sequencing and spatial transcriptomics, we constructed a cellular atlas of the aging myometrium from 186,120 cells across twenty peri- and post-menopausal women. We identified 23 myometrial cell subpopulations, including novel contractile capillary, venous capillary, immune-modulated fibroblasts, and nervous system regulatory fibroblasts. Myometrial aging leads to fewer contractile capillary cells, a reduced level of ion channel expression in smooth muscle cells, and impaired gene expression in endothelial, smooth muscle, fibroblast, perivascular, and immune cells. We observed altered myometrial cell-to-cell communication as an aging hallmark associated with the loss of 25/229 signaling pathways, including those related to angiogenesis, tissue repair, contractility, immunity, and nervous system regulation. These insights may contribute to a better understanding of the complications faced by older women during pregnancy and labor.

cell biology↗

bioRxiv2022

Brewpitopes: a pipeline to refine B-cell epitope predictions during public health emergencies

The application of B-cell epitope identification for the development of therapeutic antibodies is well established but consuming in terms of time and resources. For this reason, in the last few years, the immunoinformatic community has developed several computational predictive tools. While relatively successful, most of these tools only use a few properties of the candidate region to determine their likelihood of being a true B-cell epitope. However, this likelihood is influenced by a wide variety of protein features, including the presence of glycosylated residues in the neighbourhood of the candidate epitope, the subcellular location of the protein region or the three-dimensional information about their surface accessibility in the parental protein. In this study we created Brewpitopes, an integrative pipeline to curate computational predictions of B-cell epitopes by accounting for all the aforementioned features. To this end, we implemented a set of rational filters to mimic the conditions for the in vivo antibody recognition to enrich the B-cell epitope predictions in actionable candidates. To validate Brewpitopes, we analyzed the SARS-CoV-2 proteome. In the S protein, Brewpitopes enriched the initial predictions in 5-fold on epitopes with neutralizing potential (p-value < 2e-4). Other than S protein, 4 out of 16 proteins in the proteome contain curated B-cell epitopes and hence, have also potential interest for viral neutralization, since mutational escape mainly affects the S protein. Our results demonstrate that Brewpitopes is a powerful pipeline for the rapid prediction of refined B-cell epitopes during public health emergencies. Statement of significanceWe have created Brewpitopes, a new pipeline that integrates additional important features such as glycosylation or structural accessibility, to curate B-cell epitope more likely to be functional in vivo. We have also validated Brewpitopes against SARS-CoV-2 not only for S protein but also for the entire viral proteome demonstrating that is a rapid and reliable epitope predictive tool to be implemented in present or future public health emergencies. Brewpitopes has identified 7 SARS-CoV-2 epitopes in S and epitopes allocated in 4 other proteins. Overall, offering an accurate selection of epitopes that might be scaled up to the production of new antibodies.

bioinformatics↗

bioRxiv2022

Charting the Spatial Landscape of Cancer Hallmarks

Tumors are complex ecosystems with dozens of interacting cell types. The concept of Cancer Hallmarks distills this complexity into a set of underlying principles that govern tumor growth. Here, we exploit this abstraction to explore the physical distribution of Cancer Hallmarks across 63 primary untreated tumors from 10 cancer types using spatial transcriptomics. We show that Hallmark activity is spatially organized-with 7 out of 13 Hallmarks consistently more active in cancer cells than within the non-cancerous tumor microenvironment (TME). The opposite is true for the remaining six Hallmarks. Additionally, we discovered that genomic distance between tumor subclones correlates with differences in Cancer Hallmark activity, even leading to clone-Hallmark specialization in some cases. Finally, we demonstrate interdependent relationships between Cancer Hallmarks at the junctions of TME and cancer compartments. In conclusion, including the spatial dimension, particularly through the lens of Cancer Hallmarks, can improve our understanding of tumor ecology. SignificanceWe explored Cancer Hallmarks in 63 primary untreated tumors from 10 cancer types using spatial transcriptomics. This study unveiled spatial patterns in Hallmark activity, with some being more active in cancer cells and others in the non-cancerous tumor environment. Genomic distance impacted Hallmark activity, and we identified interdependencies at the TME-cancer junctions, improving our understanding of tumor ecology.

cancer biology↗

bioRxiv2021

A structural biology community assessment of AlphaFold 2 applications

Most proteins fold into 3D structures that determine how they function and orchestrate the biological processes of the cell. Recent developments in computational methods have led to protein structure predictions that have reached the accuracy of experimentally determined models. While this has been independently verified, the implementation of these methods across structural biology applications remains to be tested. Here, we evaluate the use of AlphaFold 2 (AF2) predictions in the study of characteristic structural elements; the impact of missense variants; function and ligand binding site predictions; modelling of interactions; and modelling of experimental structural data. For 11 proteomes, an average of 25% additional residues can be confidently modelled when compared to homology modelling, identifying structural features rarely seen in the PDB. AF2-based predictions of protein disorder and protein complexes surpass state-of-the-art tools and AF2 models can be used across diverse applications equally well compared to experimentally determined structures, when the confidence metrics are critically considered. In summary, we find that these advances are likely to have a transformative impact in structural biology and broader life science research.

biophysics↗

bioRxiv2021

The structural coverage of the human proteome before and after AlphaFold

The protein structure field is experiencing a revolution. From the increased throughput of techniques to determine experimental structures, to developments such as cryo-EM that allow us to find the structures of large protein complexes or, more recently, the development of artificial intelligence tools, such as AlphaFold, that can predict with high accuracy the folding of proteins for which the availability of homology templates is limited. Here we quantify the effect of the recently released AlphaFold database of protein structural models in our knowledge on human proteins. Our results indicate that our current baseline for structural coverage of 48%, considering experimentally-derived or template-based homology models, elevates up to 76% when including AlphaFold predictions. At the same time the fraction of dark proteome is reduced from 26% to just 10% when AlphaFold models are considered. Furthermore, although the coverage of disease-associated genes and mutations was near complete before AlphaFold release (69% of Clinvar pathogenic mutations and 88% of oncogenic mutations), AlphaFold models still provide an additional coverage of 3% to 13% of these critically important sets of biomedical genes and mutations. Finally, we show how the contribution of AlphaFold models to the structural coverage of non-human organisms, including important pathogenic bacteria, is significantly larger than that of the human proteome. Overall, our results show that the sequence-structure gap of human proteins has almost disappeared, an outstanding success of direct consequences for the knowledge on the human genome and the derived medical applications.

bioinformatics↗

bioRxiv2020

The landscape of interactions between cancer polygenic risk scores and somatic alterations in cancer cells

Over the last 15 years we have identified hundreds of inherited variants that increase the risk of developing cancer. Polygenic risk scores (PRS) summarize the genetic risk of each individual by accounting for the unique combination of risk alleles in their genome. So far, most studies of PRS in cancer have focused on their predictive value: i.e. to what extent the PRS can predict which individuals will develop a particular cancer type. In parallel, for most cancers, we have identified several subtypes based on their somatic molecular properties. However, little is known about the relationship between the somatic molecular subtypes of cancer and PRS and it is possible that PRS preferentially predict specific cancer subtypes. Since cancer subtypes can have very different outcomes, treatment options and molecular vulnerabilities, answering this question is very important to understand the consequences that widespread PRS use would have in which tumors are detected early. Here we used data from The Cancer Genome Atlas (TCGA) to study the correlation between PRS for different forms of cancer and the landscape of somatic alterations in the tumors developed by each patient. We first validated the predictive power of 8 different PRS in TCGA and describe how PRS for some cancer types are associated with specific molecular subtypes or somatic cancer driver events. Our results highlight important questions that could improve the predictive power of PRS and that need to be answered before their widespread clinical implementation.

genomics↗

bioRxiv2020

Germline genetic contribution to the immune landscape of cancer

The role of germline genetics in shaping the tumor immune landscape is largely unknown. Using genotypes from >9,000 individuals in The Cancer Genome Atlas, we investigated the association of common and rare variants with 139 well-defined immune traits. Our analysis of common variants identified 10 immune traits with significant heritability estimates, and an additional 23 with suggestive heritability, including estimates of T-cell subset abundance and interferon signaling. We performed genome-wide association on the 33 heritable traits and identified 23 genome-wide significant loci associated with at least one immune trait, including SNPs in the IFIH1 locus previously associated with several autoimmune diseases. We also found significant associations between immune traits and pathogenic or likely-pathogenic rare variants in BRCA1 and in genes functionally linked to telomere stabilization, and Wnt/Beta-catenin signaling. We conclude that germline genetic variants significantly impact the composition and functional orientation of the tumor immune microenvironment.

immunology↗

bioRxiv2020

The consequences of variant calling decisions in secondary analyses of cancer sequencing data

The analysis of cancer genomes provides fundamental information about its aetiology, the processes driving cell transformation or potential treatments. The first crucial step in the analysis of any tumor genome is the identification of somatic genetic variants that cancer cells have acquired during their evolution. For that purpose, a wide range of somatic variant callers have been developed in recent years. While there have been some efforts to benchmark somatic variant calling tools and strategies, the extent to which variant calling decisions impact the results of downstream analyses of tumor genomes remains unknown. Here we present a study to elucidate whether different variant callers (MuSE, MuTect2, SomaticSniper, VarScan2) and strategies to combine them (Consensus and Union) lead to different results in these three important downstream analyses of cancer genomics data: identification of cancer driver genes, quantification of mutational signatures and detection of clinically actionable variants. To this end, we tested how the results of these three analyses varied depending on the somatic mutation caller in five different projects from The Cancer Genome Atlas (TCGA). Our results show that variant calling decisions have a significant impact on these downstream analyses, creating important differences in driver genes identification and mutational processes attribution among variant call sets, as well as in the detection of clinically actionable targets. More importantly, it seems that Consensus, a very widely used strategy by the research community, is not the optimal strategy, as it can lead to the loss of some cancer driver genes and actionable mutations. On the other hand, the Union seems to be a legit strategy for some downstream analyses with a robust performance overall. Contact: eduard.porta@bsc.es; alfonso.valencia@bsc.es

bioinformatics↗

bioRxiv2018

Cancer3D 2.0: interactive analysis of 3D patterns of cancer mutations in cancer subsets

Our knowledge of cancer genomics exploded in last several years, providing us with detailed knowledge of genetic alterations in almost all cancer types. Analysis of this data gave us new insights into molecular aspects of cancer, most important being the amazing diversity of molecular abnormalities in individual cancers. The most important question in cancer research today is how to classify this diversity to identify subtypes that are most relevant for treatment and outcome prediction for individual patients. The Cancer3D database at http://www.cancer3d.org gives an open and user-friendly way to analyze cancer missense mutations in the context of structures of proteins they are found in and in relation to patients clinical data. This approach allows users to find novel candidate driver regions for specific subgroups, that often cannot be found when similar analyses are done on the whole gene level and for large, diverse cohorts. Interactive interface allows user to visualize the distribution of mutations is subgroups defined by cancer type and stage, gender and age brackets, patients ethnicity, or vice versa find dominant cancer type, gender or age groups for specific three-dimensional mutation patterns.

bioinformatics↗