Lu Jun, Lal Anita, Merriman Barry, Nelson Stan, Riggins Gregory
Department of Biostatistics and Bioinformatics, Duke University Medical Center, Durham, NC 27710, USA.
Genomics. 2004 Oct;84(4):631-6. doi: 10.1016/j.ygeno.2004.06.014.
A comparison study of short SAGE versus GeneChip and long SAGE was conducted to determine if data were interchangeable between the techniques. Although SAGE and Affymetrix chip expression levels showed a significant correlation using the set of genes for which there was reliable and unambiguous mapping from tag-to-gene identification, correlation coefficients were low (0.51 for the Spearman's correlation coefficient with p < 0.0001). A quantitative analysis comparing the use of 10-bp SAGE tags to 17-bp SAGE tags indicated that the short SAGE technology was more efficient at identifying differentially expressed tags. However, 4-7% of the 10-bp tags were redundancies that could have been resolved by long SAGE and 10-20% of the short SAGE tags had no obvious match to currently annotated human transcript sequences.
进行了一项短SAGE与基因芯片及长SAGE的比较研究,以确定这些技术之间的数据是否可互换。尽管使用从标签到基因识别有可靠且明确映射的基因集时,SAGE和Affymetrix芯片的表达水平显示出显著相关性,但相关系数较低(Spearman相关系数为0.51,p < 0.0001)。一项比较10碱基对SAGE标签与17碱基对SAGE标签使用情况的定量分析表明,短SAGE技术在识别差异表达标签方面更有效。然而,10碱基对标签中有4 - 7%是冗余的,可通过长SAGE解决,且10 - 20%的短SAGE标签与当前注释的人类转录本序列无明显匹配。