Periodic Reporting for period 3 - GLOMICAVE (Global Omic Data Integration on Animal, Vegetal and Environment Sectors)
Berichtszeitraum: 2023-05-01 bis 2024-04-30
Massive increases in analytical throughput together with reductions in costs have enabled multi-omics studies to be routinely performed at a scale not previously imagined. These data yield unprecedented views of intracellular mechanisms and allow the extraction of meaningful representations of signalling pathways and protein interactions networks, helpful in achieving an increased understanding of such intricate biochemical processes. Although each omics platform allows a particular comprehensive molecular survey for a given phenotype, the cross-talk between multiple molecular layers cannot be properly assessed by a reductionist approach that analyses each omics layer in isolation.
GLOMICAVE project addresses the need for building systems that allow scaling analytical processes of primary data and supporting downstream large-scale omics experiments, by maximizing the utility of pre-existing massive omics datasets to increase the understanding of biological systems as a whole, going beyond current existing datadriven tools in the biomedical sector.
The main objective of GLOMICAVE is to develop a cloud-based genotype to phenotype platform – relying on Big Data Analytics (BDA) and Artificial Intelligence (AI) techniques – and using large-scale publicly available and experimental omics datasets, enhanced with an automatic processing of scientific literature.
Current limitations / gaps that are preventing this better genotype-phenotype validation include:
i. The lack of integration of the knowledge in scientific literature and databases in an automated fashion into current genotype-phenotype association solutions.
ii. The heterogeneity of data generated by the different omics platforms preventing a homogeneous fusion and integration of data for effective computational genotype-phenotype association.
iii. The high rate of metabolites that remain unidentified in untargeted metabolomics analysis, preventing a large-scale integration of these metabolites onto metabolic networks.
iv. The poor interpretability of the results by different computational genotype-phenotype models.
v. Non-biomedical sectors lack of holistic omics science-based integration tools to solve industrial needs.
GLOMICAVE project addresses the need for building systems that allow scaling analytical processes of primary data and supporting downstream large-scale omics experiments, by maximizing the utility of pre-existing massive omics datasets to increase the understanding of biological systems as a whole, going beyond current existing datadriven tools in the biomedical sector.
The main objective of GLOMICAVE is to develop a cloud-based genotype to phenotype platform – relying on Big Data Analytics (BDA) and Artificial Intelligence (AI) techniques – and using large-scale publicly available and experimental omics datasets, enhanced with an automatic processing of scientific literature.
Current limitations / gaps that are preventing this better genotype-phenotype validation include:
i. The lack of integration of the knowledge in scientific literature and databases in an automated fashion into current genotype-phenotype association solutions.
ii. The heterogeneity of data generated by the different omics platforms preventing a homogeneous fusion and integration of data for effective computational genotype-phenotype association.
iii. The high rate of metabolites that remain unidentified in untargeted metabolomics analysis, preventing a large-scale integration of these metabolites onto metabolic networks.
iv. The poor interpretability of the results by different computational genotype-phenotype models.
v. Non-biomedical sectors lack of holistic omics science-based integration tools to solve industrial needs.
A pre-operational consolidated version of GLOMICAVE platform has been released. This version represents a significant step forward in the project, providing an integrated and demo version of the platform and installation instructions.
Users can interact with the knowledge graph through a user-friendly front-end query assistant, supported by optimized back-end strategies for near real-time query responses. The components of front-end and back-end can be easily integrated by the design of the micro service. This focuses on resources effectiveness and resistance to defects. By improving performance of graphical visualization techniques enables smooth rendering and interaction with large-scale pathway networks. In addition, the internal architecture has been updated to ensure effective data processing supported by high storage infrastructure in a scalable environment. The integration of data from various selected repositories has progressed, alongside to the collection of more scientific literature for information extraction and to refine the scientific literature mining model. The software infrastructure has been significantly upgraded to improve the ability to automatically download scientific materials, analyse data, and create detailed knowledge graphs. Moreover, efforts have been continued on the optimization and refinement of an autoML pipeline, the dysregulated pathway analysis script, and a programmatic ML-based function aimed at enhancing the data annotation module. Validation efforts across the six case studies have continued, demonstrating promising results.
The project's commitment to broader collaboration and results dissemination is demonstrated by its participation in standardization initiatives. Additionally, a final use and dissemination plan was developed to identify final potential exploitable results and how to disseminate the project results to the target audience and on innovation management. Finally, ELSI's governance has been completed emphasising the interest to strict compliance and adaptability to legislative changes, and providing a summary of the most relevant legal developments since the project's starts. In line with ethical and social responsibility, the initiative advances inclusive innovation and sustainable development, adhering to the principles of Responsible Research and Innovation (RRI).
Users can interact with the knowledge graph through a user-friendly front-end query assistant, supported by optimized back-end strategies for near real-time query responses. The components of front-end and back-end can be easily integrated by the design of the micro service. This focuses on resources effectiveness and resistance to defects. By improving performance of graphical visualization techniques enables smooth rendering and interaction with large-scale pathway networks. In addition, the internal architecture has been updated to ensure effective data processing supported by high storage infrastructure in a scalable environment. The integration of data from various selected repositories has progressed, alongside to the collection of more scientific literature for information extraction and to refine the scientific literature mining model. The software infrastructure has been significantly upgraded to improve the ability to automatically download scientific materials, analyse data, and create detailed knowledge graphs. Moreover, efforts have been continued on the optimization and refinement of an autoML pipeline, the dysregulated pathway analysis script, and a programmatic ML-based function aimed at enhancing the data annotation module. Validation efforts across the six case studies have continued, demonstrating promising results.
The project's commitment to broader collaboration and results dissemination is demonstrated by its participation in standardization initiatives. Additionally, a final use and dissemination plan was developed to identify final potential exploitable results and how to disseminate the project results to the target audience and on innovation management. Finally, ELSI's governance has been completed emphasising the interest to strict compliance and adaptability to legislative changes, and providing a summary of the most relevant legal developments since the project's starts. In line with ethical and social responsibility, the initiative advances inclusive innovation and sustainable development, adhering to the principles of Responsible Research and Innovation (RRI).
In 2024 and beyond, the impact of the GLOMICAVE project will continue to be significant in several ways.GLOMICAVE maintains its primary objective to revolutionize the analysis of genotype-phenotype relationship by comprehensively analysing and interpreting the complexity of interactions across multiple sectors. Using six different case studies covering breeding, plant breeding, and bioremediation, this platform has demonstrated its effectiveness in biomarker research, with innovative methodology and empowering researchers and end-users.
The project's multi-omics approach, coupled with the integration of unstructured bibliographic and biological data, improves the reliability of scientific conclusions. Advanced computational methods enable seamless integration of multi omics datasets, facilitating hypothesis-driven exploratory analyses. This integration unlocks the hidden relationships between biological molecules (genes, proteins, or metabolites).
The GLOMICAVE platform provides researchers with advanced tools to integrate data, analysis and predictive phenotype modelling in various industrial applications. Practical user platform interfaces make it easier to interact with multi-omic data and integrate other knowledge and analysis.
GLOMICAVE uses technology and software solutions to support high -performance research and optimize integration and data analysis processes. Machine learning methods, especially deep neural networks (DNNS), are used to analyse large datasets that occur because of multi-omics measurements, that provide effective and accurate data interpretation. Overall, GLOMICAVE's contributions help to the understanding of genotype-phenotype relationships, contributing to the development of predictive models, and generate valuable information in various industries. Through the integration of omics data, advanced analytics, and cross-disciplinary collaboration,
GLOMICAVE continues to drive innovation and impact scientific research and industrial applications. GLOMICAVE platform could be further evolved and improved by, for instance, considering the directionally of dysregulated genes/metabolites. In the sense, it might be known if a pathway is up or down regulated in the conditions compared. Besides, to consider the biological link between genes and metabolites. Hence, you can see if a certain metabolic gene dysregulation alters the metabolic profile of its target metabolites. This integration, that it is quite complex, would be very informative and valuable from a biologically point of view.
The environmental and economic impact of platform development and results of GLOMICAVE have been also assessed, foreseen benefits, environmental and economic beneifts of using GLOMICAVE platform in the selected case studies are in general wide, but a foreseen reduction of impact can be derived from the used of the platform.
The project's multi-omics approach, coupled with the integration of unstructured bibliographic and biological data, improves the reliability of scientific conclusions. Advanced computational methods enable seamless integration of multi omics datasets, facilitating hypothesis-driven exploratory analyses. This integration unlocks the hidden relationships between biological molecules (genes, proteins, or metabolites).
The GLOMICAVE platform provides researchers with advanced tools to integrate data, analysis and predictive phenotype modelling in various industrial applications. Practical user platform interfaces make it easier to interact with multi-omic data and integrate other knowledge and analysis.
GLOMICAVE uses technology and software solutions to support high -performance research and optimize integration and data analysis processes. Machine learning methods, especially deep neural networks (DNNS), are used to analyse large datasets that occur because of multi-omics measurements, that provide effective and accurate data interpretation. Overall, GLOMICAVE's contributions help to the understanding of genotype-phenotype relationships, contributing to the development of predictive models, and generate valuable information in various industries. Through the integration of omics data, advanced analytics, and cross-disciplinary collaboration,
GLOMICAVE continues to drive innovation and impact scientific research and industrial applications. GLOMICAVE platform could be further evolved and improved by, for instance, considering the directionally of dysregulated genes/metabolites. In the sense, it might be known if a pathway is up or down regulated in the conditions compared. Besides, to consider the biological link between genes and metabolites. Hence, you can see if a certain metabolic gene dysregulation alters the metabolic profile of its target metabolites. This integration, that it is quite complex, would be very informative and valuable from a biologically point of view.
The environmental and economic impact of platform development and results of GLOMICAVE have been also assessed, foreseen benefits, environmental and economic beneifts of using GLOMICAVE platform in the selected case studies are in general wide, but a foreseen reduction of impact can be derived from the used of the platform.