Technological and scientific advances have pushed the Life sciences into the BIG data era. The increased volume and heterogeneity of available data challenges scientists to master techniques for comprehensive data analysis. Extracting meaningful (biological) information from large datasets is increasingly becoming a challenge in all fields of life sciences. Thus, the ability to select and deploy analysis tools and algorithms has become an indispensable skill for all researchers. In this course, we aim to introduce participants to techniques for comprehensive data analysis of large, heterogeneous datasets and extract relevant information for elucidating biological design principles. The course is modular and focuses on data generation, mining, analysis, data integration, and visualization.
This course is aimed at anyone working with big datasets in the life sciences who has an interest in learning more about tools and possibilities for big data analysis. The course is introductory.
Required knowledge: Basic knowledge of statistics.
Prerequisite: Experience of computer programming would be useful but is not mandatory. Computer practicals will require using R. An R tutorial will be sent in advance to ensure that all participants have the required level.
The aim of this course is to introduce participants to techniques for comprehensive data analysis of big data and to integrate heterogeneous data sets in order to extract relevant information for elucidating the living system. The course is modular and focuses on data generation, mining, analysis, data integration, and visualization.
Preliminary Programme topics
Day 1 - Monday 25 October
- Linked data retrieval: querying biological resources using RDF and SPARQL.
- Brief introduction to Big data theoretical concepts: volume, variability, veracity, velocity
- Introduction to the Resource Description Framework (RDF) data format: triples, data types and objects
- Practical: Federated queries: advanced mining of (multiple) biological and biochemical database (such as UniProt, Reactome or EBI) using SPARQL
- Data reduction using multivariate statistics: PCA, theory and practical aspects
- Variance-Covariance matrix and correlations
- Data decomposition and PCA solution
- Methods for dimensionality assessment (statistical tests and computational approaches (permutation and cross-validation)
- Interpretation of PCA model: loading and biplots and limitations
Day 2 - Tuesday 26 October:
- Sparse approaches to PCA: sparse PCA, groupwise PCA
- Introduction to PLS regression
- Prediction and classification with PLS: Discriminant analysis Overfitting, Cross-validation and model optimization
- Sparse approaches to PLS
- Multilevel data integration using PLS and sparse PLS for different data types
Day 3 - Wednesday 27 October:
- Network reconstruction: Associations networks based on information theoretical methods and correlation analysis
- Brief introduction to Bayesian networks
- R tools and practical on network reconstruction exploration
- Data visualization
- Introduction to Cytoscape and R to visualize omics data such as expression data
Lecturers are from the Laboratory of Systems & Synthetic Biology (SSB) and from the Division of Human Nutrition and Health (HNH) from Wageningen University & Research:
- Edoardo Saccenti, PhD (SSB)
- Maria Suarez Diez, PhD (SSB)
- Jasper Koehorst, PhD (SSB)
- Guido Hooiveld, PhD (HNH)
Dr Edoardo Saccenti is an expert on multivariate data analysis. His research focuses on reduction and modelling techniques for large biological data sets using random matrix theory and (sparse) component approaches.
Dr Maria Suarez-Diez has extensive experience in reverse engineering of regulatory networks, metabolic modelling and the combination thereof to gain systems level understanding of the living system. She has also extensive experience in the integration of heterogeneous data sets and in the use of semantic web technologies for data integration and sharing in the life sciences.
Dr Jasper Koehorst has ample experience on infrastructure development for big data applications for bacterial genomics. He is an expert on semantic data integration and on the use of semantic resources in the life sciences
Dr Guido Hooiveld has extensive experience in the use of high throughput, information-dense omics technologies in combination with experiments in suitable models (e.g. cell, animal, and human models). He is an expert in the biological interpretation of complex nutrigenomics datasets and on integration of multi-omics data sets with functional measurements to extract novel biological insights into the molecular effects of nutrients.
Date & duration
The course will be 3 days, from 25 - 27 October 2021.
The study load of this course is 0.9 ECTS credits.
The course language will be English.
For more information on the course content please contact Dr Maria Suarez Diez (firstname.lastname@example.org) or Dr Edoardo Saccenti (email@example.com)
For organisational matters please contact Cornelia van Bree-Evers (firstname.lastname@example.org).
Registration & course fee
The course fee (which includes materials, coffee/tea during breaks, lunches and one dinner but does not cover accommodation) depends on the participant's affiliation:
|PhD candidates affiliated with VLAG, PE&RC, EPS, WIAS, WASS, SENSE/WIMEK
|| € 225
|| € 450
|University staff / Non -Profit
|| € 625
|Industry / For-Profit
|| € 1200
- No charge until 4 October 2021
- 25% of the course fee paid or due till 18 October 2021
- No refund after 18 October 2021
Substitutions for participants may be made at any time until the start of the course.