Project

WP1 Statistical data cleaning
Novel high throughput techniques enable measuring of metabolomics and glycomic datasets in large epidemiological studies. Various techniques have their own reproducibility, biases, precision and completeness (in terms of complete genome, glycome, metabolome). By developing specific correction and harmonization algorithms,WP1 will bridge the gap between high throughput methodology producing large amount of data and the methodological analysis to answer recurrent epidemiological questions. Lead: Genos (P11)

  • Deliverable 1.1: Algorithms for harmonization of similar omics datasets measured at different platforms (M36)
  • Deliverable 1.2: Imputation algorithms for phenotypic omics datasets (M60)
  • Milestone 4: For Glycomics: assessment of heterogeneity across platforms (M36)
  • Milestone 5: For metabolomics: assessment of heterogeneity across platforms (M36)
  • Milestone 6: Assessment of heterogeneity of metabolomics and Glycomis datasets across populations (M36)
  • Milestone 7: Methods for efficient signal extraction by using biological information (M36)

WP2 Systems approach to modelling pathways and structures of biological networks
Jointly with statisticians and machine learners pathway and network approaches will be extended to multilevel omics datasets (including genomic, transcriptomic, proteomic, metabolomics and glycomic datasets). Translating high throughput datasets into a better understanding of biological phenomena is one of the challenges of WP2. Application of the novel network and pathway methods to multilevel datasets, will improve the insight of biological processes underlying complex traits and elucidate biological pathways that determine the transition from health to disease (recurrent question 3). Lead: UNIBO (P5)

  • Deliverable 2.1: Algorithm for pathway reconstruction from experimental data and implementation in software (M36)
  • Deliverable 2.2: Report on network reconstruction algorithms combined with new network quantities and software (M60)
  • Milestone 1: Choice of pathway methods for proof of principle (M36 with WP7)
  • Milestone 8: Pathway classification maps for multi level omics data (M18)
  • Milestone 9: Methods for pathway analysis in epidemiological studies (M54)

WP3 Prediction, classification and clustering
In this work package prediction models based on multilevel omic datasets will be developed. We will collaborate with WP1 to include information on quality of measurement and to include biological information. We will collaborate with WP6 for computational aspects. Software packages will be developed for the methods constructed. This WP will provide the tools for answering recurrent question 1: Methods for identification of predicting molecular profiles (classification); and recurrent question 2: Methods for identification of subsets of patients groups (clustering). Input from WP2 will be used to investigate whether inclusion of biological information improves the accuracy of prediction methods. Lead: Pharmatics (P9)

  • Deliverable 3.1: Report on method for assessment of added value of distinct omics sources in predictions (M60)
  • Milestone 10: Methods for single omics source predictor methods (M36)
  • Milestone 11: Methods for combination of multiple omics sources for enhanced prediction (M54)
  • Milestone 12: Methods for identification of subgroup of patients with specific profiles (M54)

WP4 Meta-analysis
Jointly with THL methods will be developed to pool results based on single and multilevel omics dataset (Super-Meta) analysis across studies (recurrent question 4). To decrease false positive and false negative rates, researchers aim to replicate and combine results across studies. Because studies differ in applied technologies and genetic background, input from WP1 on statistical cleaning methods is particularly required in this work package. Lead: LUMC (P1)

  • Deliverable 4.1: Report on meta-analysis of multilevel biomarkers: Super-Meta (M60)
  • Milesonte 13: Methods for estimation of marginal parameters from family data (M36)
  • Milestone 14: Methods for meta-analysis of pathways (M36)
  • Milestone 15: Methods for meta analysis of – omic datasets (M54)

WP5 Causal inference
This work packages is dedicated to develop a method to establish causality of identified molecular marker profiles (recurrent question 5). Lead: UNIMAN (P15)

  • Deliverable 5.1: Causative interaction methods applied to consortia datasets (M60)
  • Milestone 16: Methods for mediation analysis (M36)
  • Milestone 17: Methods for causal inference of – omics biomarkers using sparse instrumental approach (M54)
  • Milestone 18: Causative interaction methods (M54)

WP6 Data integration and distributing computing
The methods developed in WP1 to WP5 will be applied to multiple high dimensional datasets available to the consortium (WP7 Proof of Principle). WP6 aims to provide central data storage, computational, and systems biology platforms for other project participants. This will include delivery of central data storage and computational facility and development of tools to minimize computational overhead in an external computer cluster, grid, or cloud environment. Omics analysis parallelization will be building upon our previous work on general data-level parallelization framework for GWA analyses (ParallABEL and BC genepack). Selected methods developed in this and other WPs will be brought to dissemination stage by using professional programming service. Further, new systems biology tools will be developed on the basis of GeneXplain platform, allowing integration of results achieved by consortia members with the domain of existing (systems) biological knowledge. Lead: PolyOmica (P12)

  • Deliverable 6.1: Production-stage implementation of parallelization engine of selected algorithms (M60)
  • Milestone 2: Choice of novel methods to be implemented in parallel framework (M36 – Together with WP2, WP3, WP4, WP5 and WP7)
  • Milestone 19: Central database and computational environment, web-portal (M18)
  • Milestone 20 : API between BCPlatforms and GenExplain (M36)
  • Milestone 21: Parallel implementation of existing ‘omics’analysis methods (M54)

WP7 Proof of principle (metabolic health)
For the proof-of-principle, we have chosen to focus on metabolic health. We will use the data on clinical variables, classical metabolic variables and existing metabolomics and glycomics data in cross sectional and prospective studies that allow a proof of principle analysis for a single epidemiological issue (metabolic health) and reiterated steps of improvement of the WP2-5 developed methodology. We will combine data from the consortia studies. We will use a large amount of omics data to: a) define healthy and unhealthy –omics profiles, and b) to increase the understanding of metabolic pathways that lead to poor metabolic health. UEDIN has extensive experience in interpreting and disseminating the results of similar projects, across a very wide range of illnesses and phenotypes in humans, including more than 350 disease-associated traits (including anthropometric, metabolic, physiological, clotting and inflammatory, endocrine, plasma protein N-glycans, IgG N-glycans, plasma ceramides, plasma sphingolipids, plasma phospholipids, plasma lipoproteins and urinary metabolites) and related diseases. Lead: UEDIN (P3)

  • Deliverable 7.1: Report on results of analysis of consortia data using MIMOmics methods (M60)
  • Milestone 3: Choice of novel methods for proof of principle (M36 – Together with WP2, WP3, WP4 and WP5)
  • Milestone 22: Pathway analysis for metabolic health (M60 – Together with WP2)
  • Milestone 23: Detection of novel biomarkers for metabolic health (M60 – Together with WP3)
  • Milestone 24: Assessment of causality of biomarkers for metabolic health (M60 – Together with WP5)

WP8 Dissemination: Health care, Industry and Academia
GeneXplain is a SME which aims to provide a comprehensive platform for bioinformatics and systems biological tools. Via their broad commercial contact network and their marketing channels to industrial partners in life sciences, the methods and tools developed during the project will be propagated. Jointly with the rest of the involved SMEs they will play a pivotal role in dissemination of the MIMomics output. The co-leader of this WP is UCAM which has a broad experience in organizing training programs for Academia. Using both traditional channels of scientific communication as well as new, electronic and Internet-based media, the consortium partners will make efforts to disseminate the methodological achievements of the project. The information overload as a general problem also in the scientific community makes it necessary to accompany high-quality scientific publications with appropriate “marketing” measures. Lead: GENEXPLAIN GMBH (P14)

  • Deliverable 8.1: Report on organized courses and symposia (M60)
  • Milestone 25: Awareness using social media organized (M36)
  • Milestone 26: Materials ready for commercial presentations (M54)
  • Milestone 27: Exploration of valorisation possibilities (M60)

WP9 Project Management
Lead: LUMC (P1)

  • Milestone 28: Project office in place (M1)
  • Milestone 29: Organizing kick-off meeting (M2)
  • Milestone 30: Management functions in place (M6)
  • Milestone 31: Website up and running (M6)