From reference genes to global mean normalization Jo Vandesompele - - PowerPoint PPT Presentation

From reference genes to global mean normalization

Jo Vandesompele professor, Ghent University co-founder and CEO, Biogazelle qPCR Symposium USA November 9, 2009 – Millbrae, CA

• utline

 what is normalization  gold standard for mRNA normalization  global mean normalization and selection of stable small RNAs for

microRNA normalization

introduction to normalization

 2 sources of variation in gene expression results

 biological variation (true fold changes)  experimentally induced variation (noise and bias)

 purpose of normalization is reduction of the experimental variation

 input quantity: RNA quantity, cDNA synthesis efficiency, …  input quality: RNA integrity, RNA purity, …

 gold standard is the use of multiple stably expressed reference genes

 which genes?  how many?  how to do the calculations?

normalization: geNorm solution

 framework for qPCR gene expression normalisation using the reference

gene concept:

 quantified errors related to the use of a single reference gene

(> 3 fold in 25% of the cases; > 6 fold in 10% of the cases)

 developed a robust algorithm for assessment of expression stability of

candidate reference genes

 proposed the geometric mean of at least 3 reference genes for

accurate and reliable normalisation

 Vandesompele et al., Genome Biology, 2002

geNorm software

 automated analysis

 ranking of candidate reference genes according to their stability  determination of how many genes are required for reliable normalization  http://medgen.ugent.be/genorm

0.003 0.006 0.021 0.023 0.056 NF4 NF1

 cancer patients survival curve

statistically more significant results

geNorm validation (I)

log rank statistics Hoebeeck et al., Int J Cancer, 2006

 mRNA haploinsufficiency measurements

accurate assessment of small expression differences

geNorm validation (II)

Hellemans et al., Nature Genetics, 2004



patient / control



3 independent experiments



95% confidence intervals

 geNorm is the de facto standard for reference gene validation and

normalization

 > 2,000 citations of our geNorm technology  > 10,000 geNorm software downloads in 100 countries

normalization using multiple stable reference genes

global mean normalization

 when a large set of genes are measured, the average expression level

reflects the input amount and could be used for normalization

 e.g. microarray based normalization

• lowess, mean ratio, …

 SAGE / NGS sequencing counts

 the set of genes must be unbiased and sufficiently large  we make use of this principle to normalize microRNA data from

experiments in which we quantify a substantial number of miRNAs (450 or 650) in a given sample

global mean normalization

 small-RNA controls

 classic normalization strategy  small nuclear RNAs, small nucleolar RNAs  18 available from Applied Biosystems

 global mean normalization

 method applied for microarray data  universal: applicable for every miRNA dataset  many datapoints needed (megaplex vs. multiplex)

 miRNAs/controls that resemble the mean

 minimal standard deviation when comparing miRNA expression with

mean ( geNorm V value, standard deviation of log transformed ratios)

 compatible with multiplex assays  need to determine mean

small RNA controls

 How ‘stable’ is the global mean compared to controls?

 geNorm analysis using controls and mean as input variables  exclusion of potentially co-regulated controls

HY3 7q36 RNU19 5q31.2 RNU24 9q34 RNU38B 1p34.1-p32 RNU43 22q13 RNU44 1q25.1 RNU48 6p21.32 RNU49 17p11.2 RNU58A 18q21 RNU58B 18q21 RNU66 1p22.1 RNU6B 10p13 U18 15q22 U47 1q25.1 U54 8q12 U75 1q25.1 Z30 17q12 RPL21 13q12.2

miRNA expression datasets

 neuroblastoma tumour samples  T-ALL samples  EVI1 deregulated leukemias  retinoblastoma tumour samples  normal tissues  normal bone marrow

0,2 0,4 0,6 0,8 1 1,2 1,4 1,6 1,8 expression stability

T-ALL geNorm ranking

geNorm ranking bone marrow pool normal tissues neuroblastoma leukemia EVI1 overexpression

20 40 60 80 100 120 50 100 150 200 250 300 not normalised stable controls mean miRNAs

neuroblastoma – removal of variation

removal of variation bone marrow pool normal tissues T-ALL leukemia EVI1 overexpression

biological validation

 MYCN binds to the mir-17-92 promoter

CpG island mir-17-92 cluster +5 kb

• 5 kb

CATGTG CATGTG CATGTG CACGTG CACGTG CATGTG CATGTG

A B C

1 2 3 4 5 6 7 8 9 10 11 12 A B C

Fold enrichment Amplicon

IMR5 WAC2

biological validation

 choice of normalization strategy influences differential miRNA expression

 Mir-17-92 expression in neuroblastoma tumours

0,5 1 1,5 2 2,5 3 3,5

stable controls mean miRNAs

biological validation

 choice of normalization strategy influences differential miRNA expression

 Mir-17-92 expression in neuroblastoma tumours

0,5 1 1,5 2 2,5 3 3,5

stable controls mean miRNAs

biological validation

 choice of normalization strategy influences differential miRNA expression

 Mir-17-92 expression in neuroblastoma tumours

0,5 1 1,5 2 2,5 3 3,5

stable controls mean miRNAs

• 7
• 6
• 5
• 4
• 3
• 2
• 1

1 2 3 fold change (MYCN amplified vs. MYCN single copy) controls mean

average FCcontrols = -0.404 average Fcmean = 0.050 average FCmiRNAs = 0.124

balanced differential expression

correlation MYCN downregulated genes – 2 normalization strategies stable miRNA control normalisation mean normalisation

strategy also works for microarray data

 each sample is measured by

RT-qPCR and microarray

 global mean normalization  standardization per method  hierarchical clustering  samples cluster by sample

(and NOT by method)

conclusions global mean normalization

 novel and powerful miRNA normalization strategy

 maximal reduction of technical noise  improved identification of differentially expressed genes  balancing of differential expression  universally applicable

• global mean
• multiple stable endogenous controls

 Mestdagh et al., Genome Biology, 2009

 most powerful, flexible and user-friendly real-time PCR data-analysis

software

 based on Ghent University’s geNorm and qBase technology  state of the art normalization procedures

• one or more classic reference genes
• global mean normalization
• expressed repeat normalization

 detection and correction of inter-run variation  dedicated error propagation  fully automated analysis; no manual interaction required  booth 19

qbasePLUS normalization

http://www.qbaseplus.com

conclusions

 proper normalization has a major impact on your results

 provides statistically more significant results  enables accurate assessment of small expression differences

 gold standard for mRNA gene expression analysis

 geNorm evaluation of candidate reference genes  geometric mean of multiple stably expressed reference genes

 global mean normalization and subsequent geNorm based selection of

reference genes that resemble the mean is a valid option when measuring a large and unbiased set of genes (e.g. all miRNAs)

acknowledgments

 miRNA

 Pieter Mestdagh (UGent)  Frank Speleman (UGent)  Applied Biosystems

 qbasePLUS

 Jan Hellemans (Biogazelle – UGent)  Stefaan Derveaux (Biogazelle – UGent)

 January 28-29, 2010

Ghent, Belgium

 www.advances-in-genomics.org