Multivariate Analysis and Visualization of Proteomic Data
Download as PPT, PDF2 likes2,890 views
The document discusses the multivariate analysis and visualization of proteomic data, highlighting the challenges of data quality assessment, statistical analysis, and methodologies for analyzing large datasets. It emphasizes the importance of understanding experimental design and using statistical tools to detect trends, classify samples, and map biochemical networks. Additionally, it provides insights into available software resources for data analysis and visualization.
Multivariate Analysis and Visualization of Proteomic Data
- 1. Multivariate Analysis and Visualization of ProteOmic Data Dmitry Grapov, PhD
- 2. State of the art facility producing massive amounts of biological data… >20-30K samples/yr >200 studies
- 3. Analysis at the ProteOmic Scale and Beyond Genomic Omic Multi-Omic Metabolomic Proteomic integration
- 4. Sample Data Analysis and Visualization Variable Quality Assessment • use replicated mesurements and/or internal standards to estimate analytical variance Statistical and Multivariate • use the experimental design to test hypotheses and/or identify trends in analytes Functional • use statistical and multivariate results to identify impacted biochemical domains Network • integrate statistical and multivariate results with the experimental design and analyte metadata Sample Variable experimental design - organism, sex, age etc. analyte description and metadata - biochemical class, mass spectra, etc.
- 5. Sample Data Analysis and Visualization Variable Quality Assessment • use replicated mesurements and/or internal standards to estimate analytical variance Statistical and Multivariate • use the experimental design to test hypotheses and/or identify trends in analytes Functional • use statistical and multivariate results to identify impacted biochemical domains Network • integrate statistical and multivariate results with the experimental design and analyte metadata Network Mapping Sample Variable experimental design - organism, sex, age etc. analyte description and metadata - biochemical class, mass spectra, etc.
- 6. Data Quality Assessment Quality metrics •Precision (replicated measurements) •Accuracy (reference samples) Common tasks •normalization •outlier detection •missing values imputation
- 7. Batch Effects Drift in >400 replicated measurements across >100 analytical batches for a single analyte Principal Component Analysis (PCA) of all analytes, showing QC sample scores Acquisition batch Abundance QCs embedded among >5,5000 samples (1:10) collected over 1.5 yrs If the biological effect size is less than the analytical variance then the experiment will incorrectly yield insignificant results
- 8. Analyte specific data quality overview Sample specific normalization can be used to estimate and remove analytical variance Raw Data Normalized Data Normalizations need to be numerically and visually validated low precision log mean %RSD high precision Samples QCs Batch Effects
- 9. Outlier Detection • 1 variable (univariate) • 2 variables (bivariate) • >2 variables (multivariate)
- 10. bivariate vs. multivariate (scatter plot) outliers? mixed up samples (PCA scores plot) Outlier Detection
- 11. Statistical and Multivariate Analyses Group 1 Statistics Multivariate Context + + = Network Mapping Ranked statistically significant differences within a a biochemical context Group 2 What analytes are different between the two groups of samples? Statistical t-Test significant differences lacking rank and context Multivariate O-PLS-DA ranked differences lacking significance and context
- 12. Statistical and Multivariate Analyses Group 1 Statistics Multivariate Context + + = Network Mapping Group 2 What analytes are different between the two groups of samples? Statistical t-Test Multivariate O-PLS-DA To see the big picture it is necessary too view the data from multiple different angles
- 13. Statistical Analysis: achieving ‘significance’ significance level (α) and power (1-β ) effect size (standardized difference in means) sample size (n) Power analyses can be used to optimize future experiments given preliminary data Example: use experimentally derived (or literature estimated) effect sizes, desired p-value (alpha) and power (beta) to calculate the optimal number of samples per group
- 14. Statistical Tests Poisson normal • Should be chosen based on the distribution (shape, type) of the (e.g. normal, negative binomial, Poisson) • Can be optimized based on data pre-treatment (e.g. NSAF, Power Law Global Error Model, PLGEM)
- 15. False Discovery Rate (FDR) Type I Error: False Positives (α) •Type II Error: False Negatives (β) •Type I risk = •1-(1-p.value)m m = number of variables tested
- 16. False Discovery Rate Adjustment FDR adjusted p-value p-value Benjamini & Hochberg (1995) (“BH”) •Accepted standard Bonferroni •Very conservative •adjusted p-value = p-value x # of tests (e.g. 0.005 x 148 = 0.74 )
- 17. Functional Analysis Identify changes or enrichment in biochemical domains • decrease • increase Nucl. Acids Res. (2008) 36 (suppl 2): W423-W426.doi: 10.1093/nar/gkn282
- 18. Functional Analysis: Enrichment Biochemical Pathway Biochemical Ontology
- 19. Common Multivariate Methods Clustering Projection Networks
- 20. Artist: Chuck Close Cluster Analysis Useful for •pattern recognition •complexity reduction Common Methods •Hierarchical •Model based •Other (k-means, k-NN, PAM, fuzzy) Linkage k-means Distribution Density
- 21. Hierarchical Clustering Similarity x x x x Dendrogram How does my metadata match my data structure?
- 22. Projection Methods single analyte all analytes The algorithm defines the position of the light source Principal Components Analysis (PCA) • unsupervised • maximize variance (X) Partial Least Squares Projection to Latent Structures (PLS) • supervised • maximize covariance (Y ~ X) James X. Li, 2009, VisuMap Tech.
- 23. Interpreting scores and loadings loadings represent how variables contribute to sample scores variables with the highest loadings have the greatest contribution to sample scores loadings scores Scores represent dis/similarities in samples based on all variables
- 24. Networks Biochemical •interaction • enrichment •etc Empirical (dependency) •correlation •partial-correlation •clustering variable 1 variable 2 variable 3
- 25. Enrichment Network Mapping of parents through children
- 27. Empirical Networks • Correlation based networks (CN) (simple, tendency to hairball) • GGM or partial correlation based networks (advanced, preference of direct over indirect relationships • *Increase in robustness with sample size 10.1007/978-1-4614-1689-0_17
- 28. Proteomic Case Study: Diabetes Markers • Small sample size (control =12, GDM =6); covariates (time of sample collection) • >600 measured colostrum proteins; ~ 300 NSAF normalized proteins retained • Multivariate classification with O-PLS-DA used to identify variables to test using PLGEM with correction for FDR • Partial-correlation protein-protein interaction network analysis
- 29. DeviumWeb https://github.com/dgrapov/DeviumWeb • visualization • statistics • clustering • PCA • O-PLS
- 30. DeviumWeb https://github.com/dgrapov/DeviumWeb • visualization • statistics • clustering • PCA • O-PLS
- 31. Software and Resources •DeviumWeb- Dynamic multivariate data analysis and visualization platform url: https://github.com/dgrapov/DeviumWeb •imDEV- Microsoft Excel add-in for multivariate analysis url: http://sourceforge.net/projects/imdev/ •MetaMapR- Network analysis tools for metabolomics url: https://github.com/dgrapov/MetaMapR •TeachingDemos- Tutorials and demonstrations •url: http://sourceforge.net/projects/teachingdemos/?source=directory •url: https://github.com/dgrapov/TeachingDemos •Data analysis case studies and Examples url: http://imdevsoftware.wordpress.com/
- 32. Questions? dgrapov@ucdavis.edu This research was supported in part by NIH 1 U24 DK097154