E5510 Biomarkers Detection from Omics Experiments

Faculty of Science
Autumn 2022
Extent and Intensity
2/0/0. 2 credit(s) (plus extra credits for completion). Type of Completion: zk (examination).
Teacher(s)
Mgr. Eva Budinská, Ph.D. (lecturer)
Mgr. Soňa Smetanová, Ph.D. (assistant)
Guaranteed by
Mgr. Eva Budinská, Ph.D.
RECETOX – Faculty of Science
Contact Person: Mgr. Eva Budinská, Ph.D.
Supplier department: RECETOX – Faculty of Science
Timetable
Fri 10:00–11:50 D29/347-RCX2
Prerequisites
Basic knowledge of molecular biology and statistics
Course Enrolment Limitations
The course is offered to students of any study field.
The capacity limit for the course is 25 student(s).
Current registration and enrolment status: enrolled: 1/25, only registered: 0/25, only registered with preference (fields directly associated with the programme): 0/25
Course objectives
The goal of this course is to introduce the main principles of analysis of the data from molecular 'omics experiments (microarrays, mass spectrometry, NGS...). A special emhpasis will be given on experimental design and result validation in biomarker detection. Typical examples will be presented with in depth discussion of issues and pitfalls of specific methods and approaches.
Learning outcomes
At the end of this course the student: - can correctly define the hypothesis of the omics experiment - has knowledge of the batch effects and other confounding factors and can minimize them - identifies a suitable method for data analysis based on the null hypothesis - has knowledge of common features of molecular omics experiments - has knowledge of the technology-specific data pre-processing steps - has knowledge of the statistical methods of groups comparison, group discovery, classification and survival analysis with emphasis on omics data analysis - uses a suitable graphical representation of the results and is able to correctly interpret them, - is able to select suitable validation set and identify the best validation method - applies the acquired knowledge in design and validation of omics experiments
Syllabus
  • 1. Introduction. „To consult the statistician after an experiment is finished, is often merely to ask him to conduct a post mortem examination. He can perhaps say what the experiment died of.“ This lecture will give a general overview of common concepts of omics data types, data pre-processing, statistical analysis and validation of the results and why it matters if we want to produce robust and reproducible results. 2. Its all about batch effects… - In this lecture we will show a number of examples on how different lab-work steps influence the downstream data analysis and introduce bias in the results. We will discuss best laboratory practices and experimental design minimizing batch effects as well as statistical techniques for batch effect removal. 3.-4. Finding differences between two or more groups. In two lectures, we will explain how the lists of differentially expressed features (genes, proteins...) between the groups of interest are created and why most of the classical statistical approaches (e.g. T-test, …) is not applicable on this type of data. The issue of multiple hypothesis testing and its correction will be discussed. Recommendations for experimental design will be given. Most common graphical and table representation of the results will be explained. 5.-6. Searching for similar patterns / groups. When the aim of the study is to find new groups in the data, based on similar molecular patterns of expression, clustering techniques are used. We will introduce the most suitable clustering approaches that are able to create robust and reproducible results. Cluster validation techniques and considerations for experimental design will be discussed. In addition, the most common graphical and table representation of the results will be explained. 7.-8. Classify-it! (How (not) to predict almost anything). In two lectures we will discuss in detail the best practices of development of prognostic and predictive classifiers based on molecular data (biomarkers) and how classification helps to identify the most relevant features (proteins, genes…) or sample outliers. Special emphasis will be given on classifier validation as well as proper experimental design. The most common graphical and table representation of the results will be explained. 9. Pathway analysis – techniques of gene set enrichment, topological pathway analysis and gene network identification will be explained. 10. Meta-analysis. This lecture will introduce techniques of meta-analysis for comparison of results from different studies. 11. Databases. Where are the omics data stored, how to upload our experimental data and what are the minimal requirements. 12. Experimental design and validation – a wrap-up lecture that will pool the information on experimental design and validation from all previous lectures and fill in the missing pieces. A complete guideline for experimental design and validation will be presented.
Literature
    recommended literature
  • FORSHED, Jenny. Experimental Design in Clinical ‘Omics Biomarker Discovery. Journal of Proteome Research. Washington: American Chemical Society, 2017, vol. 16, No 11, p. 3954-3960. ISSN 1535-3893. info
  • The MicroArray Quality Control (MAQC)-IIII study of common practices for the development and validation of microarray-based predictive models. Nature biotechnology. NEW YORK: NATURE PUBLISHING GROUP, 2010, vol. 28, No 8, p. "827"-"U109", 15 pp. ISSN 1087-0156. Available from: https://dx.doi.org/10.1038/nbt.1665. info
  • GHOSH, Debashis and Laila M POISSON. “Omics” data and levels of evidence for biomarker discovery. Genomics. United States: Elsevier Science Inc, 2009, vol. 2009, 93(1), p. 13-16. ISSN 0888-7543. info
Teaching methods
The course is taught in form of oral presentations and exercises online using a web application R-shiny. Emphasis is placed on understanding the mechanisms and concepts. Students are often asked about the currently discussed topic and encouraged to ask questions and interact with the teacher.
Assessment methods
During the lectures, students are asked about subjects of past lecture. Final assessment (at the end of semester) is by written examination. It is not a multiple choice test but a set of questions, which frequently require description, explanation or schematization of the topic. Questions have 1 - 3 points according to their difficulty. Total count is 20 and to pass at least 10 points are needed.
Language of instruction
Czech
Further Comments
Study Materials
The course is taught annually.
The course is also listed under the following terms Autumn 2023, Autumn 2024.
  • Enrolment Statistics (Autumn 2022, recent)
  • Permalink: https://is.muni.cz/course/sci/autumn2022/E5510