Division of Dermatology and Cutaneous Sciences, Center for Investigative Dermatology, 4179 Biomedical and Physical Sciences Building, College of Human Medicine, Michigan State University, East Lansing, MI 48823, USA.
Br J Dermatol. 2009 Dec;161(6):1278-87. doi: 10.1111/j.1365-2133.2009.09338.x. Epub 2009 Jun 11.
DNA microarray technology has revealed vast numbers of gene expression alterations associated with human malignancies. Assigning validity and biological significance to these changes, however, remains a considerable hurdle. Recently, microarray analysis has been applied to the study of nonmelanoma skin cancer.
To compare experimental data rigorously in order to strengthen conclusions regarding the pathogenesis of basal cell carcinoma (BCC) and squamous cell carcinoma (SCC), and to evaluate systematically the experimental and statistical parameters that may impact the degree of consensus among differentially expressed genes (DEGs) between studies.
We performed a systematic comparison of 10 studies that applied DNA microarray technology to study BCC/SCC.
A total of 1133 DEGs collectively reported across the studies were compared, and 64 DEG overlaps were found: 18 DEG overlaps in SCC vs. SCC study comparisons, 18 DEG overlaps in BCC vs. BCC study comparisons and 28 DEG overlaps in BCC vs. SCC study comparisons. We documented differences in several experimental methods that may account for the relative lack of consensus between studies, including sample type, tissue procurement/handling, microarray chip and statistical analysis. The two most dysregulated biological pathways across all studies involved genes with enzymatic and structural/adhesion functions.
DEGs that were found to overlap across two or more studies and biological pathways with the largest representation of DEGs across studies may be particularly relevant to disease pathogenesis and serve as targets for future therapy. In future work, more consistent experimental methods across laboratories may improve the validity of reported DEGs and strengthen conclusions drawn from microarray data.
DNA 微阵列技术揭示了与人类恶性肿瘤相关的大量基因表达改变。然而，赋予这些变化有效性和生物学意义仍然是一个相当大的障碍。最近，微阵列分析已应用于非黑色素瘤皮肤癌的研究。
为了加强对基底细胞癌（BCC）和鳞状细胞癌（SCC）发病机制的结论，严格比较实验数据，并系统评估可能影响研究之间差异表达基因（DEG）一致性程度的实验和统计参数。
我们对应用 DNA 微阵列技术研究 BCC/SCC 的 10 项研究进行了系统比较。
总共比较了 10 项研究中报告的 1133 个共同的 DEG，发现了 64 个 DEG 重叠：18 个 SCC 与 SCC 研究比较中的 DEG 重叠，18 个 BCC 与 BCC 研究比较中的 DEG 重叠和 28 个 BCC 与 SCC 研究比较中的 DEG 重叠。我们记录了可能导致研究之间相对缺乏共识的几个实验方法的差异，包括样本类型、组织采集/处理、微阵列芯片和统计分析。所有研究中两个最失调的生物学途径都涉及具有酶和结构/粘附功能的基因。
在两个或更多研究中发现重叠的 DEG 和在研究中具有最大 DEG 代表性的生物途径可能与疾病发病机制特别相关，并可作为未来治疗的靶点。在未来的工作中，实验室之间更一致的实验方法可能会提高报告的 DEG 的有效性，并加强从微阵列数据中得出的结论。
