Analysing data from Illumina BeadArrays

Transcription

1 The bead Analysing data from Illumina BeadArrays Each silica bead is 3 microns in diameter Matt Ritchie Department of Oncology University of Cambridge, UK 4th September ,000 copies of same probe sequence are covalently attached to each bead for hybridisation & decoding Can have more than one bead for a particular gene * Bead preparation and array production Beads in wells Bead pools produced containing 384 to 4,000 bead types Wells created in either fibre optic bundle (hexagon) or chip (rectangle) Beads self-assemble into wells to form randomly arranged array of beads Average of 30 beads of each type (adds robustness) Each array produced separately SAM - Sentrix Array Matrix Decoding process Beads get assigned random locations on each array. Need to know what type every bead is Decoding is achieved by series of sequential hybridisations* 0 0 Each bead type defined by unique DNA sequence that is recognised by a complementary decoder Process is highly effective (error rate < x 0-4 per bead) * Gunderson et al. Decoding randomly ordered DNA arrays. Genome Research, May arrays processed in parallel each array contains ~ 500 bead types

2 BeadChip Advantages Random placement of beads High number of within array replicates - robustness High throughput studies possible (eg HapMap) RefSeq BeadChip 8 arrays per chip strip = array 4,000 bead types from RefSeq database x 30 reps on each array Whole Genome 6 arrays per chip: strips = array 48,000 bead types on each array (4,000 RefSeq + 4,000 supplementary) Expression, genotyping and methylation variants possible Lower cost Data analysis array.locs files Scanning (BeadScan) Image analysis (BeadScan) Summarisation (BeadScan) Quality assessment (BeadStudio) Normalisation (BeadStudio) Downstream analysis (BeadStudio) TIFFs.txt files.idat files summary.txt Raw data Illumina s scanning software (BeadScan) produces files (.idat,.locs etc) in proprietary format which are read by Illumina s BeadStudio software However, with modifications to BeadScan, you can obtain standard (readable) files for each array on a SAM or strip on a BeadChip - Text file giving the identity and location of each individual bead (~ 50,000 rows for each SAM ~. million for each strip/array on a BeadChip) [required] -TIFF images [optional] We refer to the text and TIFF files as the bead-level data for an array Example: Bead-level text files TIFF images ProbeID Bead Centre Background corrected intensity Information for all* beads on an array Can use bead centres to re-calculate intensities *Sometimes outliers or non-decoded beads are removed

3 Image analysis Foreground signal = average over 9 pixels for each bead Background calculated using a 7 x 7 window around each bead Local background = average of the 5 minimum pixels ( ) within each window Background correction carried out by BeadScan: Foreground - Background For two-colour spotted arrays, background correction in this way can be disastrous! Spike-in data set Mouse 6 version BeadChips Arrays include 33 spike probes Each array hybridised with the same RNA sample + spike 000pM 300pM 00pM 30pM 0pM 3pM x 4 pm 0.3pM 0.pM 0.03pM 0.0pM 0pM x 4 Concentration series Raw data Foreground Background log (intensity) 3pM vs pm pm vs 0.3pM 3

4 Concentration series Quality assessment - image plot scanning problem - values down one edge censored at 0 4-7% of beads affected - successfully detected as outliers Question: How many outliers can Illumina s summarisation method handle before we need to think about removing an array? Simulation of corrupted data Background normalisation bias variance % outliers simulated 3pM vs pm Use of variability in DE analysis Summarise data from replicate beads on the log -scale (mean and variance) Use inverse of bead variances as weights in differential expression (DE) analysis log-odds Assess DE with and without (treat all observations equally) weights 4

5 Recommendations Illumina s local background correction and summarisation methods perform well We do not currently recommend Illumina s background normalisation (now referred to as subtract background option in BeadStudio - not to be confused with local background subtraction refereed to above, which is not optional) We do not recommend exporting gene summary data from BeadStudio, as the results are averaged over (sometimes) misannotated probes. We prefer probe summary output. Access to bead-level data enables - more detailed quality assessment - calculation of means and variances on the appropriate scale The beadarray package from Bioconductor can process bead-level or summary-level data in R. We have used limma for DE analysis. Current and future work improved annotation (for next Bioconductor release) automatic artefact detection and removal tool (BASH) exploring data from two-colour technologies apply crlmm to Illumina Infinium II genotyping data Acknowledgements Tavaré Lab Mark Dunning Andy Lynch Nuno Barbosa-Morais Jonathan Cairns Mike Smith Simon Tavaré Genomics Core James Hadfield Michelle Osborne Illumina (San Diego) Semyon Kruglyak Gary Nunn UCSD (San Diego) Roman Sasik WEHI (Melbourne) Wei Shi Gordon Smyth References. KL Gunderson, S Kruglyak, et al. Decoding randomly ordered DNA arrays. Genome Res, 4(5): , 004. K Kuhn, SC Baker, et al. A novel, high-performance random array platform for quantitative gene expression profiling. Genome Res, 4(): , M Barnes, J Freudenberg, et al. Experimental comparison and cross-validation of the Affymetrix and Illumina gene expression analysis platforms. Nucleic Acids Res, 33(8): , MJ Dunning, ML Smith, et al. beadarray: R classes and methods for Illumina beadbased arrays, Bioinformatics, 3(6):83-4, MJ Dunning, NL Barbosa-Morais, et al. Statistical issues in the analysis of Illumina data. BMC Bioinformatics, 9:85, MJ Dunning, ME Ritchie, et al. Spike-in validation of an Illumina-specific variancestabilizing transformation. BMC Research Notes, :8, Illumina probe reannotation: Accessing bead-level data Bead-level text and tiff files can usually be obtained by modifying the following lines in the settings.xml file used by BeadScan (the entries below are typically be set to their opposites, i.e. true -> false, and false -> true): <SavePerBeadFiles>true</SavePerBeadFiles> <SaveTextFiles>true</SaveTextFiles> <ExcludeOutliers>false</ExcludeOutliers> <IncludeXY>true</IncludeXY> 5