Spatial bias

Spatial variation is observed even in the best laboratories [1-5]. Spatial bias can be plotted in form of heat maps. The following heat maps were selected at random, and show six dual-channel microarrays. The microarrays were printed and hybridised according to the protocols given by the UK Drosophila core microarray facility web site (FlyChip). The x/y-coordinates represent location on the microarray, whilst the colouring and contours indicate the difference between the Cy5 and Cy3 channels. These six examples show that spot location has a direct impact on the gene expression ratios and hence supports the need for randomisation of the spot layout to facilitate correction of spatial biases by normalisation [7-10].

First example	Second example
Two patches of high gene expression ratios towards the centre of the array and decreased ratios towards the edges, especially the right edge. Top and bottom are roughly equivalent to the central region.	Gradual increase in the measured gene expression ratios from bottom-to-top. The rate of increase varies between the left and right, leading to a second gradient across the centre of the array from right-to-left.

Third example	Forth example
Low gene expression ratios at the centre, top and bottom of the array. Higher ratios towards the left and right edges.	Low ratios at the left and right edges that appears to increase towards the centre and then again at the top of the microarray.

Fifth example	Sixth example
High ratios at the bottoma and top-left corner, low ratios for one patch on the left edge and the entire top-right corner.	Reasonably even ratios across most of the microarray, but with slightly higher ratios in the top left corner.

Further reading

The following publications discuss are a small sub-set of the papers that describe systematic biases that affect microarray experiments and the need for randomised spot layouts to facilitate correction of these biases by normalisation.

1. Wernisch L, Kendall SL, Soneji S, Wietzorrek A, Parish T, Hinds J, Butcher PD, Stoker NG: Analysis of whole-genome microarray replicates using mixed models. Bioinformatics 2003, 19(1):53-61
2. Qian J, Kluger Y, Yu H, Gerstein M: Identification and correction of spurious spatial correlations in microarray data. Biotechniques 2003, 35(1):42-44, 46, 48
3. Yang YH, Buckley MJ, Dudoit S, Speed TP: Comparison of methods for image analysis on cDNA microarray data. J Comp Graph Stat 2002, 11(1):108-136
4. Yang YH, Dudoit S, Luu P, Lin DM, Peng V, Ngai J, Speed TP: Normalization for cDNA microarray data: a robust composite method addressing single and multiple slide systematic variation. Nucleic Acids Res 2002, 30(4):e15
5. Futschik ME, Crompton T: OLIN: optimized normalization, visualization and quality testing of two-channel microarray data. Bioinformatics 2005, 21(8):1724-1726
6. Kreil DP, Russell RR: There is no silver bullet--a guide to low-level data transforms and normalisation methods for microarray data. Brief Bioinform 2005, 6(1):86-97
7. Kerr MK, Churchill GA: Statistical design and the analysis of gene expression microarray data. Genet Res 2001, 77(2):123-128
8. Churchill GA: Fundamentals of experimental design for cDNA microarrays. Nat Genet 2002, 32 Suppl:490-495
9. Brodsky L, Leontovich A, Shtutman M, Feinstein E: Identification and handling of artifactual gene expression profiles emerging in microarray hybridization experiments. Nucleic Acids Res 2004, 32(4):e46
10. Le Meur N, Lamirault G, Bihouee A, Steenman M, Bedrine-Ferran H, Teusan R, Ramstein G, Leger JJ: A dynamic, web-accessible resource to process raw microarray scan data into consolidated gene expression values: importance of replication. Nucleic Acids Res 2004, 32(18):5349-5358