Reply to "Comment on: What genes are differentially expressed in individuals with schizophrenia? A systematic review".
Reply to "Comment on: What genes are differentially expressed in individuals with schizophrenia? A systematic review".
复制标题
回复“评论：精神分裂症个体中哪些基因差异表达？系统评价”。
DOI：
10.1038/s41380-022-01821-2
复制
发表时间：
2023
影响因子：
11
通讯作者：
Almasy,Laura
中科院分区：
文献类型：
--
作者：
Merikangas,AlisonK;Almasy,Laura
我们感谢有机会回应霍夫曼和他的同事对我们最近对精神分裂症基因表达研究的系统回顾的评论[1]。我们同意这些作者在评论中提出的所有问题，包括样本量和统计能力的极端重要性，对可能混淆关联的潜在技术和样本特定因素的控制，以及需要区分与精神分裂症的病因有关的基因表达与疾病的后果。霍夫曼和他的同事的评论只关注死后大脑中的基因表达，我们的综述包括对血液和其他组织的研究，如下所述。正如霍夫曼和同事所指出的，力量是由样本大小驱动的。我们将我们的文献回顾追溯到2000年，以努力捕捉到精神分裂症最早的全基因组基因表达研究。与最近的研究相比，这些早期研究的样本量确实较小。然而，当我们考虑到三项或更多研究报告的基因数量时，排除较小的研究不会改善以这种方式评估的结果的重叠。霍夫曼等人。建议一种包括所有文本的荟萃分析方法，而不仅仅是那些被个别研究报告为重要的文本，将是一种更强有力的策略。我们在这一点上是完全一致的。荟萃分析是我们最初的目标[2]，但进行这样的荟萃分析所需的数据并不容易获得。因此，我们强调了基因表达研究结果的公共储存库的必要性，类似于现有的全基因组关联研究(GWAS)。正如我们所指出的，“公开共享的基因表达结果储存库的创建将促进实验室小组之间信息的广泛共享，并允许更容易的荟萃分析”。
We appreciate the opportunity to respond to Hoffman and colleagues’ commentary on our recent systematic review of gene expression studies in schizophrenia [1]. We agree with these authors on all of the issues raised in their comments, including the critical importance of sample size and statistical power, control for potential technical and sample-specific factors that may confound associations, and the need to distinguish between gene expression that is involved in the etiology of schizophrenia as opposed to being a consequence of the disease. Whereas Hoffman and colleagues’ comments focused solely on gene expression in postmortem brain, our review included studies in blood and other tissues, as described below.As Hoffman and colleagues note, power is driven by sample size. We extended our literature review back to the year 2000 in an effort to capture the earliest genome-wide gene expression studies in schizophrenia. These early studies do indeed have smaller sample sizes than more recent efforts. However, as we considered the number of genes reported by three or more studies, excluding smaller studies would not improve overlap in findings assessed in this manner. Hoffman et al. suggest that a meta-analytical approach including all transcripts, rather than only those reported as significant by individual studies, would be a stronger strategy. We are in complete agreement on this point. A meta-analysis had been our initial aim [2], but the data needed to conduct such meta-analyses were not readily available. Therefore, we highlighted the need for public repositories of gene expression study results, similar to those already in existence for genomewide association studies (GWAS). As we noted “The creation of publicly shared gene expression result repositories would facilitate wide-spread sharing of information between lab groups and allow for much easier meta-analyses”.