In many cancers, mechanisms of gene regulation can be severely altered. Identification of deregulated genes, which do not follow the regulation processes that exist between transcription factors and their target genes, is of importance to better understand the development of the disease. We propose a methodology to detect deregulation mechanisms with a particular focus on cancer subtypes. This strategy is based on the comparison between tumoral and healthy cells. First, we use gene expression data from healthy cells to infer a reference gene regulatory network. Then, we compare it with gene expression levels in tumor samples to detect deregulated target genes. We finally measure the ability of each transcription factor to explain these deregulations. We apply our method on a public bladder cancer data set derived from The Cancer Genome Atlas project and confirm that it captures hallmarks of cancer subtypes. We also show that it enables the discovery of new potential biomarkers.

中文翻译：

---

确定特定于癌症亚型的失调机制

在许多癌症中，基因调控机制可能会发生严重改变。识别不遵循转录因子与其靶基因之间存在的调节过程的失调基因对于更好地了解疾病的发展具有重要意义。我们提出了一种检测放松管制机制的方法，特别关注癌症亚型。该策略基于肿瘤细胞和健康细胞之间的比较。首先，我们使用来自健康细胞的基因表达数据来推断参考基因调控网络。然后，我们将其与肿瘤样本中的基因表达水平进行比较，以检测失调的靶基因。我们最终测量了每个转录因子解释这些失调的能力。我们将我们的方法应用于源自癌症基因组图谱项目的公共膀胱癌数据集，并确认它捕获了癌症亚型的标志。我们还表明，它能够发现新的潜在生物标志物。