Lercher Martin J, Hurst Laurence D
Department of Biology and Biochemistry, University of Bath, Bath BA2 7AY, UK.
J Mol Biol. 2006 Jun 9;359(3):825-31. doi: 10.1016/j.jmb.2006.03.051. Epub 2006 Apr 6.
Analysis in yeast of the relationship between a gene's genomic position and its expression profile, derived from chip array data, suggests that both closely linked genes and genes spaced at regular intervals show correlated expression profiles. Unfortunately, yeast arrays are often printed in genomic order. The above results may hence reflect little more than known spatial biases within arrays. To circumvent this problem, we analyse spatially unbiased expression data derived from a large Northern blot study. We find that local domains of co-expressed genes range up to 30 genes (100 kb), and are thus much larger than previously considered. There is, by contrast, no evidence for periodicity of co-expression in yeast. We likewise find no convincing evidence for periodicity in the human or mouse genome. Further, analysis of yeast transcription factor binding data sets suggests that there is currently no statistical evidence for chromosomal periodicity of co-regulation.
通过酵母对基因的基因组位置与其表达谱之间关系（源自芯片阵列数据）的分析表明，紧密连锁的基因以及以规则间隔排列的基因都呈现出相关的表达谱。不幸的是，酵母阵列通常按基因组顺序打印。因此，上述结果可能仅仅反映了阵列内已知的空间偏差。为了规避这个问题，我们分析了来自一项大型Northern印迹研究的空间无偏表达数据。我们发现共表达基因的局部结构域范围可达30个基因（100 kb），因此比之前认为的要大得多。相比之下，没有证据表明酵母中存在共表达的周期性。我们同样没有在人类或小鼠基因组中找到令人信服的周期性证据。此外，对酵母转录因子结合数据集的分析表明，目前没有统计证据支持共调控的染色体周期性。
