Discovery of new deregulated miRNAs in gingivo buccal carcinoma using Group Benjamini Hochberg method: a commentary on “A quest for miRNA bio-marker: a track back approach from gingivo buccal cancer to two different types of precancers”

This formal comment is in response to “A quest for miRNA bio-marker: a track back approach from gingivo buccal cancer to two different types of precancers” written by De Sarkar and colleagues in 2014. The above-mentioned paper found seven miRNAs to be significantly deregulated in 18 gingivo-buccal cancer samples. However, they suspected more miRNAs to be deregulated based on their exploratory statistical analysis. To control the false discovery rate (FDR), the authors used the Benjamini Hochberg (BH) method, which does not leverage any available biological information on the miRNAs. In this work, we show that some specialized versions of the BH method, which can exploit positional information on the miRNAs, can lead to seven more discoveries with this data. Specifically, we group the closely located miRNAs, and use the group Benjamini Hochberg (GBH) methods (Hu et al., 2010), which reportedly have more statistical power than the BH method (Liu et al., 2019). The whole transcriptome analysis of Sing et al. (2017) and previous literature on the miRNAs suggest that most of the newly discovered miRNAs play a role in oncogenesis. In particular, the newly discovered miRNAs include hsa-miR-1 and hsa-miR-21-5p, whose cancer-related activities are well-established. Our findings indicate that incorporating the GBH method into suitable microarray studies may potentially enhance scientific discoveries via the exploitation of additional biological information.