Reproductive phasi RNA s in grasses are compositionally distinct from other classes of small RNA s

Summary Little is known about the characteristics and function of reproductive phased, secondary, small interfering RNA s (phasi RNA s) in the Poaceae, despite the availability of significant genomic resources, experimental data, and a growing number of computational tools. We utilized machine‐learning methods to identify sequence‐based and positional features that distinguish phasi RNA s in rice and maize from other small RNA s ( sRNA s). We developed Random Forest classifiers that can distinguish reproductive phasi RNA s from other sRNA s in complex sets of sequencing data, utilizing sequence‐based (k‐mers) and features describing position‐specific sequence biases. The classification performance attained is > 80% in accuracy, sensitivity, specificity, and positive predicted value. Feature selection identified important features in both ends of phasi RNA s. We demonstrated that phasi RNA s have strand specificity and position‐specific nucleotide biases potentially influencing AGO sorting; we also predicted targets to infer functions of phasi RNA s, and computationally assessed their sequence characteristics relative to other sRNA s. Our results demonstrate that machine‐learning methods effectively identify phasi RNA s despite the lack of characteristic features typically present in precursor loci of other small RNA s, such as sequence conservation or structural motifs. The 5′‐end features we identified provide insights into AGO –phasi RNA interactions. We describe a hypothetical model of competition for AGO loading between phasi RNA s of different nucleotide compositions.