Laboratoire de Biométrie et Biologie Evolutive, Université de Lyon, CNRS, UMR5558, 69622 Villeurbanne, France.
INRIA Lyon Centre, 69100 Villeurbanne, France.
Genes (Basel). 2023 Mar 7;14(3):664. doi: 10.3390/genes14030664.
By pairing to messenger RNAs (mRNAs for short), microRNAs (miRNAs) regulate gene expression in animals and plants. Accurately identifying which mRNAs interact with a given miRNA and the precise location of the interaction sites is crucial to reaching a more complete view of the regulatory network of an organism. Only a few experimental approaches, however, allow the identification of both within a single experiment. Computational predictions of miRNA-mRNA interactions thus remain generally the first step used, despite their drawback of a high rate of false-positive predictions. The major computational approaches available rely on a diversity of features, among which anchoring the miRNA seed and measuring mRNA accessibility are the key ones, with the first being universally used, while the use of the second remains controversial. Revisiting the importance of each is the aim of this paper, which uses Cross-Linking, Ligation, And Sequencing of Hybrids (CLASH) datasets to achieve this goal. Contrary to what might be expected, the results are more ambiguous regarding the use of the seed match as a feature, while accessibility appears to be a feature worth considering, indicating that, at least under some conditions, it may favour anchoring by miRNAs.
通过与信使 RNA（简称 mRNA）配对，microRNAs（miRNAs）在动植物中调节基因表达。准确识别与特定 miRNA 相互作用的 mRNAs 以及相互作用位点的精确位置，对于更全面地了解生物体的调控网络至关重要。然而，只有少数实验方法可以在单个实验中同时识别这两者。因此，尽管计算预测 miRNA-mRNA 相互作用存在高假阳性预测率的缺点，但仍然是普遍采用的第一步。现有的主要计算方法依赖于多种特征，其中锚定 miRNA 种子和测量 mRNA 可及性是关键特征，前者被普遍使用，而后者的使用仍存在争议。本文旨在重新审视每个特征的重要性，为此使用交联、连接和杂交测序 (CLASH) 数据集来实现这一目标。出乎意料的是，关于将种子匹配用作特征的使用结果更加模棱两可，而可及性似乎是一个值得考虑的特征，表明至少在某些条件下，它可能有利于 miRNA 的锚定。
