Pasar al contenido principal

NOVEL METHODOLOGIES FOR MULTI-OMICS INTEGRATIVE ANALYSIS WITH CUTTING-EDGE SEQUENCING TECHNOLOGIES

PID2023-152976NB-I00

•
Nombre agencia financiadora Agencia Estatal de Investigación
Acrónimo agencia financiadora AEI
Programa Programa Estatal para Impulsar la Investigación Científico-Técnica y su Transferencia
Subprograma Subprograma Estatal de Generación de Conocimiento
Convocatoria Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Año convocatoria 2023
Plan Estatal 2021-2023 Plan Estatal de Investigación Científica y Técnica y de Innovación
Centro beneficiario AGENCIA ESTATAL CONSEJO SUPERIOR DE INVESTIGACIONES CIENTIFICAS (CSIC)

Resultados relacionados

Resultados totales (Incluyendo duplicados): 8
Encontrada(s) 1 página(s)

RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

MOREshiny: a user-friendly application for the inference of phenotype-specific multi-omic regulatory networks

  • Aguerralde-Martin, Maider|||0000-0001-9266-8689
  • Tarazona, Sonia|||0000-0001-5346-1407
  • Moldovan-Moldovan, Roxana Andreea
  • Verdú, María
[EN] Deciphering phenotype-specific regulatory mechanisms is key to understanding the molecular basis of complex diseases and traits. However, constructing multi-omic regulatory networks (MO-RNs) is challenging, as it requires integrating heterogeneous omics data, incorporating biological context, and detecting regulatory mechanisms that vary across conditions. The R package MORE (Multi-Omics REgulation) addresses these challenges by applying robust statistical models to infer phenotype-specific regulatory networks from multi-omics data. However, the use of MORE typically requires programming expertise, limiting its accessibility to non-specialist users. To democratize access to advanced multi-omics modeling tools, we present MOREshiny, an interactive web application built on Shiny that extends the module of pathway enrichment analysis and automatically guides the choice of statistical methods. MOREshiny enables users to upload multi-omic data, configure their models, and interpret results through a user-friendly interface-without the need for coding skills. MOREshiny also allows users to download MORE results for their later exploration and study. To demonstrate the utility of MOREshiny, we showcase its functionalities on a multi-omic ovarian cancer dataset to understand regulatory differences between patients who did or did not require chemotherapy., This work was supported by the Scientific Foundation of the Spanish Association Against Cancer [PERME224336TARA] and the Instituto de Salud Carlos III [AC22/00058] (co-funded by the European Union as part of the Next Generation EU programme and the Recovery and Resilience Mechanism, MRR), both under the frame of ERA PerMed [ERAPERMED2022-141 OVA-PDM]; and the Spanish Ministry of Science, Innovation and Universities [PID2023-152976NB-I00, PREP2023-001818]. Funding for open access charge: Universitat Politecnica de Valencia.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Tesis doctoral (DoctoralThesis)

Interpretable Survival Analysis for High-Dimensional and Multi-Block data

  • Salguero García, Pedro
[ES] El panorama médico actual se define por el crecimiento exponencial de los datos, desde los registros clínicos digitalizados hasta la secuenciación ómica de alta resolución. Si bien esta información promete revolucionar nuestra comprensión de las enfermedades, su vasto volumen introduce barreras analíticas y metodológicas significativas. Mi investigación comenzó a través de la participación en la Iniciativa E-Yellow Submarine, una colaboración europea centrada en la mortalidad por COVID-19 y las diferencias de supervivencia basadas en el sexo. El desafío inmediato no fue la falta de datos, sino la heterogeneidad de los mismos: los hospitales participantes albergaban datos clínicos invaluables, pero sus conjuntos de datos estaban aislados por formatos incompatibles, métodos heterogéneos de adquisición de variables y poblaciones de pacientes diferentes. Para superar este obstáculo, la primera contribución de esta tesis fue el desarrollo de ATHAR, una herramienta semiautomática diseñada específicamente para estandarizar, unificar y explorar metódicamente estos flujos de datos dispares en una única base de datos armonizada y analizable. Aunque exitoso, este paso inicial de armonización expuso de inmediato una profunda barrera analítica. Específicamente, la inclusión de un único conjunto de datos de alta dimensionalidad (AD) en la colaboración reveló las limitaciones metodológicas fundamentales del análisis de supervivencia tradicional. Los marcos clásicos, como el modelo Cox, demostraron ser inestables y propensos al sobreajuste al enfrentarse a la naturaleza de datos de AD. Esta brecha y la necesidad de herramientas que gestionen la complejidad estadística al tiempo que ofrecen una clara interpretabilidad biológica motivaron directamente la segunda contribución principal de esta tesis: Coxmos. Coxmos fue diseñado como un paquete de R de código abierto para confrontar esta frontera metodológica. Coxmos sinergia el modelo Cox con las capacidades de reducción de dimensionalidad y selección de variables de las metodologías de Mínimos Cuadrados Parciales (PLS). Esta metodología fue diseñada no solo para la predicción, sino para crear modelos totalmente interpretables para el análisis de supervivencia en datos de AD y multibloque (MB). Nuestra extensa validación demuestra que Coxmos no solo supera los métodos clásicos, sino que también supera consistentemente a las técnicas de aprendizaje automático de última generación en escenarios de AD y MB. En resumen, esta tesis documenta una trayectoria científica completa, que abarca desde el desafío práctico de armonizar datos clínicos brutos hasta la creación de un marco estadístico para el análisis de supervivencia de AD y MB. Ambas contribuciones, ATHAR y Coxmos, establecen un flujo de trabajo integral de extremo a extremo. Estos recursos equipan colectivamente a la comunidad científica para transformar datos complejos en conocimiento robusto e interpretable, acelerando directamente el descubrimiento de biomarcadores pronósticos y avanzando en la medicina de precisión., [CA] El panorama mèdic actual es definix pel creixement exponencial de les dades, des dels registres clínics digitalitzats fins a la seqüenciació òmica d'alta resolució. Si bé esta informació promet revolucionar la nostra comprensió de les malalties, el seu vasto volum introduïx barreres analítiques i metodològiques significatives. La meua investigació va començar a través de la participació en la Iniciativa E-Yellow Submarine, una col·laboració europea centrada en la mortalitat per COVID-19 i les diferències de supervivència basades en el sexe. El desafiament immediat no va ser la falta de dades, sinó l'heterogeneïtat de les mateixes: els hospitals participants albergaven dades clíniques invaluables, però els seus conjunts de dades estaven aïllats per formats incompatibles, mètodes heterogenis d'adquisició de variables i poblacions de pacients diferents. Per a superar este obstacle, la primera contribució d'esta tesi va ser el desenvolupament d' ATHAR, una ferramenta semiautomàtica dissenyada específicament per a estandarditzar, unificar i explorar metòdicament estos fluxos de dades dispars en una única base de dades harmonitzada i analitzable. Encara que reeixit, este pas inicial d'harmonització va exposar immediatament una profunda barrera analítica. Específicament, la inclusió d'un únic conjunt de dades d'alta dimensionalitat (AD) en la col·laboració va revelar les limitacions metodològiques fonamentals de l'anàlisi de supervivència tradicional. Els marcs clàssics, com el model Cox, van demostrar ser inestables i propensos al sobreajust en enfrontar-se a la naturalesa de dades d'AD. Esta bretxa i la necessitat de ferramentes que gestionen la complexitat estadística al mateix temps que oferixen una clara interpretabilitat biològica van motivar directament la segona contribució principal d'esta tesi: Coxmos. Coxmos va ser dissenyat com un paquet d'R de codi obert per a confrontar esta frontera metodològica. Coxmos sinergitza el model Cox amb les capacitats de reducció de dimensionalitat i selecció de variables de les metodologies de Mínims Quadrats Parcials (PLS). Esta metodologia va ser dissenyada no només per a la predicció, sinó per a crear models totalment interpretables per a l'anàlisi de supervivència en dades d'AD i multibloc (MB). La nostra extensa validació demostra que Coxmos no sols supera els mètodes clàssics, sinó que també supera consistentment a les tècniques d'aprenentatge automàtic de última generació en escenaris d'AD i MB. En resum, esta tesi documenta una trajectòria científica completa, que abasta des del desafiament pràctic d'harmonitzar dades clíniques brutes fins a la creació d'un marc estadístic per a l'anàlisi de supervivència d'AD i MB. Ambdós contribucions, ATHAR i Coxmos, establixen un flux de treball integral d'extrem a extrem. Estos recursos equipen col·lectivament a la comunitat científica per a transformar dades complexes en coneixement robust i interpretable, accelerant directament el descobriment de biomarcadors pronòstics i avançant en la medicina de precisió., [EN] The contemporary medical landscape is defined by the exponential growth of data, from digitized clinical records to high-resolution omic sequencing. While this information promises to revolutionize our understanding of disease, its vast volume introduces significant analytical and methodological barriers. My research began through involvement in the E-Yellow Submarine Initiative, a European collaboration focused on COVID-19 mortality and sex-based survival differences. The immediate challenge was not a lack of data but data heterogeneity: participating hospitals housed invaluable clinical data, yet their datasets were isolated by incompatible formats, heterogeneous variable acquisition methods, and differing patient populations. To overcome this obstacle, the first contribution of this thesis was the development of ATHAR, a semi-automatic pipeline specifically engineered to methodically standardize, unify, and explore these disparate data streams into a single, analyzable harmonized database. Although successful, this initial harmonization step immediately exposed a profound analytical barrier. Specifically, the inclusion of a single HD dataset in the collaboration revealed the fundamental methodological limitations of traditional survival analysis. Classical frameworks, such as Cox model, proved unstable and prone to overfitting when faced with the HD nature of the data. This gap and the need for tools that manage statistical complexity while delivering clear biological interpretability directly motivated the second major contribution of this thesis: Coxmos. The resulting Coxmos framework was engineered as an open-source R package to confront this methodological frontier. Coxmos synergizes the Cox model with the dimensionality reduction and variable selection capabilities of PLS. This framework was designed not merely for prediction, but to create fully interpretable models for HD and MB survival data. Our extensive validation demonstrates that Coxmos not only surpasses classical methods but also consistently outperforms state-of-the-art ML techniques in HD and MB scenarios. In summary, this thesis documents a complete scientific arc, ranging from the practical challenge of harmonizing raw clinical data to the creation of a statistical framework for HD and MB survival analysis. The dual contributions, ATHAR and Coxmos, establish a comprehensive, end-to-end workflow. These resources collectively equip the scientific community to transform complex data into robust, interpretable knowledge, directly accelerating the discovery of prognostic biomarkers and advancing the clinical realization of precision medicine., The research presented in this thesis, including the development of ATHAR and the Coxmos package, was made possible by the funding and support of several grants. The thesis was funded by the ACIF/2019/081 grant awarded by the Generalitat Valenciana, the PREVENTIC (2021-073-1) project awarded by Agència Valenciana de la Innovació, the PID2020-119537RB-I00 grant awarded by the Spanish Ministerio de Ciencia e Innovación, the MIMARISCA-2 (PI2021- 10) project funded by the UPV and the project "Novel methodologies for multi-omics integrative analysis with cutting-edge sequencing technologies"(PID2023-152976NB-I00) funded by the Spanish Ministerio de Ciencia, Innovación y Universidades.
RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia
Publicación Artículo científico (JournalArticle)

Coxmos: interpretable survival models for high-dimensional and multi-omic data

  • Salguero, Pedro
  • Buendía-Galera, Anabel
  • Tarazona, Sonia|||0000-0001-5346-1407
[EN] Background Survival analysis in high-dimensional (HD) and multi-block (MB) settings, such as omic and multi-omic studies, poses major methodological challenges due to multicollinearity, low events-per-variable ratios, and limited model interpretability. While Cox Proportional Hazards models remain widely used, their applicability is restricted in HD scenarios. Machine learning approaches can handle such complexity but often lack interpretability. Survival models based on Partial Least Squares (PLS) offer an attractive compromise, yet existing implementations provide limited support for model optimization, evaluation, and interpretation.Results To overcome these limitations, we introduce Coxmos, a CRAN R package that integrates adapted Cox regression models with HD-adapted variable selection, and optimized PLS-based approaches tailored for HD and MB data. Coxmos also provides validation, comparison, interpretation, and visualization tools. Benchmarking on a diverse set of clinical, single-omic, and multi-omic cancer datasets - including breast (BRCA), head and neck (HNSC), and ovarian (HGSOC) cancer - demonstrated that Coxmos outperformed other state-of-the-art machine learning methods, while enhancing interpretability. We illustrated the package's functionalities through two case studies (HGSOC and HNSC), showing how Coxmos facilitates model selection, validation, and biological interpretation in a complex HD and MB survival analysis.Conclusions Coxmos provides a robust, flexible, and interpretable solution for survival analysis in HD and MB data. By combining optimized PLS-based methodologies with rigorous evaluation and rich visualization tools, Coxmos enables reliable survival prediction and facilitates the identification of relevant risk and protective factors in complex biomedical studies. The package fills an important methodological gap and represents a valuable resource for reproducible and interpretable survival modeling in the big data scenario., This work was supported by the Scientific Foundation of the Spanish Association Against Cancer through the project PERME224336TARA and by Instituto de Salud Carlos III through the project AC22/00058 (co-funded by the European Union as part of the Next Generation EU programme and the Recovery and Resilience Mechanism, MRR), both projects under the frame of ERA PerMed (ERAPERMED2022-141 - OVA-PDM). This work was also funded by Generalitat Valenciana (ACIF/2019/081), Spanish MICIN (PID2020-119537RB-100), and Spanish MICIU (PID2023-152976NB-I00).
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Notable challenges posed by long-read sequencing for the study of transcriptional diversity and genome annotation

  • Monzó, Carolina
  • Frankish, Adam
  • Conesa, Ana
Long-read sequencing (LRS) technologies have revolutionized transcriptomic research by enabling the comprehensive sequencing of full-length transcripts. Using these technologies, researchers have reported tens of thousands of novel transcripts, even in well-annotated genomes, while developing new algorithms and experimental approaches to handle the noisy data. The Long-read RNA-seq Genome Annotation Assessment Project community effort benchmarked LRS methods in transcriptomics and validated many novel, lowly expressed, often times sample-specific transcripts identified by long reads. These molecules represent deviations of the major transcriptional program that were overlooked by short-read sequencing methods but are now captured by the full-length, single-molecule approach. This Perspective discusses the challenges and opportunities associated with LRS’ capacity to unravel this fraction of the transcriptome, in terms of both transcriptome biology and genome annotation. For transcriptome biology, we need to develop novel experimental and computational methods to effectively differentiate technology errors from rare but real molecules. For genome annotation, we must agree on the strategy to capture molecular variability while still defining reference annotations that are useful for the genomics community., This work was supported by a grant from the National Institutes of Health (1R21HG011280-01), the Spanish Ministry of Science Innovation and Universities (MCIU) (PID2020-119537RB-I00 and PID2023-152976NB-I00), and the European Union's programme Horizon Europe under a Marie Skłodowska-Curie grant (101149931). The computations were performed on the high performance computing cluster Garnatxa at the Institute for Integrative Systems Biology (I2SysBio), I2SysBio is a joint research institute of the University of Valencia (UV), and Spanish National Research Council (CSIC)., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Quality assessment of long read data in multisample lrRNA-seq experiments with SQANTI-reads

  • Keil, Netanya
  • Monzó, Carolina
  • McIntyre, Lauren M.
  • Conesa, Ana
SQANTI-reads leverages SQANTI3, a tool for the analysis of the quality of transcript models, to develop a read-level quality control framework for replicated long-read RNA-seq experiments. The number and distribution of reads, as well as the number and distribution of unique junction chains (transcript splicing patterns), in SQANTI3 structural categories are informative of raw data quality. Multisample visualizations of QC metrics are presented by experimental design factors to identify outliers. We introduce new metrics for (1) the identification of potentially under-annotated genes and putative novel transcripts and for (2) quantifying variation in junction donors and acceptors. We applied SQANTI-reads to two different data sets, a Drosophila developmental experiment and a multiplatform data set from the LRGASP project and demonstrate that the tool effectively reveals the impact of read coverage on data quality, and readily identifies strong and weak splicing sites., This work was supported in part by a grant from the National Institutes of Health (1R21HG011280-01), the Spanish Ministry of Science Innovation and Universities (MCIU) (PID2020-1195 37RB-I00 and PID2023-152976NB-I00), the European Union’s programme Horizon Europe under the Marie Skłodowska-Curie Actions postdoctoral fellowship to C.M. (101149931). This work is also supported in part bya grant fromthe NationalCancerInstitute (NCI P01 CA214091), the National Institute of General Medical Sciences (NIGMS GM137430), the University of Florida Department of Molecular Genetics and Microbiology, the University of Florida Genetics Institute, the University of Florida Cancer Center, and the University of Florida Research Computing Center (www.rc.ufl.edu) and the Latin American and Caribbean Scholars award to N.K. Part of the computations were performed on the high-performance computing cluster Garnatxa at the Institute for Integrative Systems Biology (I2SysBio). I2SysBio is a joint research center formed by the University of Valencia (UV) and the Spanish National Research Council (CSIC). We acknowledge Ashley Myrick for help with some of the initial CV plots and Knife Bankole for the initial coding of the junction hash. Alison Morse prepared all samples and libraries for the Drosophila experiment, ran all initial QC analyses, and recalled all the bases for ONT data. We acknowledge Rolf Renne for his support., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Long reads decipher genomes and transcriptomes and offer novel insights into biology and diseases

  • Conesa, Ana
  • Hoischen, Alexander
  • Sedlazeck, Fritz J.
Editorial., A.H. was supported by a ZonMW (The Netherlands Organization for Health Research and Development) Vici grant no. 09150182310053. F.J.S. was supported by the National Institutes of Health (1UG3NS132105-01 and 1U01HG011758-01). A.C. was supported by the Spanish Ministry of Science and Innovation grant no. PID2023-152976NB-I00., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

TUSCO: benchmarking transcriptome reconstruction with endogenous single-isoform controls

  • Liu, Tianyuan
  • Paniagua, Alejandro
  • Jetzinger, Fabian
  • Ferrández-Peral, Luis
  • Frankish, Adam
  • Conesa, Ana
[Data availability] Publicly available LRS datasets used in this study include the LRGASP Consortium data for human WTC11 iPSC and mouse ES cells, accessible at the Synapse repository under accession syn25007472/wiki/608702. The RNA degradation datasets are deposited in the European Nucleotide Archive (ENA) under accession PRJEB53210, with supplementary FASTQ data for post-mortem brain samples available from the European Genome-Phenome Archive (EGA) under accession EGAS00001006542. The Replicated Mouse Brain and Kidney dataset, generated in-house, has been submitted in its entirety to the ENA under accessions PRJEB85167 (secondary accession ERP168594, “Sequencing of mice brains using Iso-Seq”) and PRJEB94912 (secondary accession ERP177675, “Sequencing of mice kidneys using Iso-Seq”). For benchmarking, the universal and tissue-specific TUSCO gene sets for human and mouse can be freely obtained at https://github.com/ConesaLab/SQANTI3 and at the dedicated TUSCO benchmarking framework portal https://tusco.uv.es. An example of the TUSCO report is available at: https://github.com/ConesaLab/tusco-paper/blob/main/example/tusco_output/tusco_report.html. The example input data and full TUSCO output files are provided in the example/ directory of the same repository. Source data are provided with this paper. All data are also available from the corresponding author upon request., Long-read sequencing (LRS) platforms, such as Oxford Nanopore and Pacific Biosciences, enable comprehensive transcriptome analysis but face challenges such as sequencing errors, sample quality variability, and library preparation biases. Current benchmarking approaches address these issues insufficiently: BUSCO assesses transcriptome completeness using conserved single-copy orthologous genes but can misinterpret alternative splicing as gene duplications, while SIRV spike-ins and ERCCs oversimplify real sample complexity, neglecting RNA degradation and RNA-extraction artifacts, thus inflating performance metrics. Simulation algorithms are limited in their ability to recapitulate the complexity of real samples. To overcome these limitations, we introduce the Transcriptome Universal Single-isoform COntrol (TUSCO) benchmarking framework, centered on a curated TUSCO gene set of genes lacking alternative isoforms that can be confidently treated as an internal ground truth. The TUSCO evaluation quantifies precision by identifying reconstructed transcripts that deviate from reference annotations and quantifies sensitivity by verifying detection completeness in human and mouse samples. Masking TUSCO gene set transcripts and replacing them with modified splice variants in the annotation creates a TUSCO-novel challenge that assesses reconstruction of the true, now-unannotated isoforms. Our validation demonstrates that TUSCO metrics provide accurate and reliable benchmarking without external controls, significantly improving quality control standards for transcriptome reconstruction using LRS., Computational analyses were performed on the high-performance computing cluster Garnatxa at the Institute for Integrative Systems Biology (I2SysBio). I2SysBio is a joint research centre formed by University of Valencia (UV) and Spanish National Research Council (CSIC). T.L., F.J., and A.C. disclose support for the research of this work from the European Union Marie Skłodowska-Curie Actions Doctoral Network project LongTREC [HORIZON-MSCA-2021-DN-01, grant agreement 101072892]. A.P. discloses support for the research of this work from the Spanish MICINN [FPU21/01597]. A.C. and L.F.P. disclose support for the research of this work from the National Human Genome Research Institute of the National Institutes of Health [1R21HG011280-01]. A.C. discloses additional support from MICIU/AEI [PID2023-152976NB-I00]. A.F. discloses support for the research of this work from the National Human Genome Research Institute of the National Institutes of Health [U24HG007234, U24HG011451], Wellcome Trust [WT222155/Z/20/Z], and the European Molecular Biology Laboratory., Peer reviewed
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publicación Artículo científico (JournalArticle)

Handling biological replicates in long-read RNA sequencing data by joining or not joining

  • Jetzinger, Fabian
  • Paniagua, Alejandro
  • Cormack, Stanley
  • Mestre-Tomás, Jorge
  • Morante-Redolat, José Manuel
  • Fariñas, Isabel
  • Ferrández-Peral, Luis
  • Götz, Stefan
  • Monzó, Carolina
  • Conesa, Ana
[Data availability] Mice brain and kidney data generated using PacBio, ONT and Illumina sequencing were submitted to the European Nucleotide Archive (ENA; https://www.ebi.ac.uk/ena/browser/home) and are accessible under accession numbers PRJEB851674 [https://www.ebi.ac.uk/ena/browser/view/PRJEB85167] (mouse brain), PRJEB9491235 [https://www.ebi.ac.uk/ena/browser/view/PRJEB94912] (mouse kidney). The human R10.4.1 nanopore cDNA sequencing data analyzed in this study are publicly available from the NCBI Sequence Read Archive under BioProject accession PRJNA1169789, originally generated by DeBruyne et al.33. Specifically, we used the four whole-transcriptome cDNA replicates of the SH-SY5Y neuroblastoma cell line: SRR31732191, SRR31732198, SRR31732209, and SRR31732210. Source data are provided with this paper., [Code availability] To facilitate future evaluation of J&C vs C&J strategies with additional datasets, transcript reconstruction tools and their configurations, we provide a nextflow46 pipeline that enables systematic evaluation of the multi-sample combination strategies. It is available at https://github.com/ConesaLab/join_and_call_paper, This project has received funding from the European Union’s programme Horizon Europe under the Marie Skłodowska-Curie grant agreement Number 101072892 and 101149931, the Spanish MICINN (PID2020-119537RB-I00 and FPU21/01597), MICIU/AEI (PID2023-152976NB-I00) and the National Institutes of Health (1R21HG011280-01)., Peer reviewed