Accurate Plant MicroRNA Prediction Can Be Achieved Using Sequence Motif Features

Abstract

MicroRNAs (miRNAs) are short (~21 nt) nucleotide sequences that are either co-transcribed during the production of mRNA or are organized in intergenic regions transcribed by RNA polymerase II. In animals, Drosha, and in plants DCL1 recognize pre-miRNAs which set themselves apart by their characteristic stem loop (hairpin) structure. This structure appears important for their recognition during the process of maturation leading to functioning mature miRNAs. A large body of research is available for computational pre-miRNA detection in animals, but less within the plant kingdom. For the prediction of pre-miRNAs, usually machine learning approaches are employed. Therefore, it is necessary to convert the pre-miRNAs into a set of features that can be calculated and many such features have been described. We here select a subset of the previously described features and add sequence motifs as new features. The resulting model which we called Motif-miRNAPred was tested on known pre-miRNAs listed in miRBase and its accuracy was compared to existing approaches in the field. With an accuracy of 99.95% for the generalized plant model, it distinguishes itself from previously published results which reach an average accuracy between 74% and 98%. We believe that our approach is useful for prediction of pre-miRNAs in plants without per species adjustment.

Computational methods for the ab initio identification of novel microRNA in plants: a systematic review

“…Eleven studies used accuracy as a performance measure and 10 of those studies achieved accuracies above 90%. Even though the reported performances are not directly comparable, the highest accuracy of 99.48% was reported by Yousef et al, (2016). Considering the results presented by each study, all of them performed well and therefore, are seemingly reliable.…”

“…As almost all reported miRNAs are found in the non-coding regions of the genome, these sequences are assumed as pseudo miRNA data (Xuan et al, 2011). Guan et al (2011), Koh & Kim (2017), Xiao et al (2011), Yousef, Allmer & Khalifa (2015) and Yousef et al (2016) used the negative datasets from previous studies which were already available. Yousef, Allmer & Khalifa (2015) discusses a one-class classifier for plant miRNAs where they only used the positive data set.…”

“…Different studies have been conducted to show the impact of different sets of features. Some methods show that thermodynamic features (Yao et al, 2016) are better while another reports that sequential features (Yousef et al, 2016) are better. However, there is no concrete answer or common theme since there aren't many studies comparing different feature types for plant miRNA prediction.…”

Categorization of species based on their microRNAs employing sequence motifs, information-theoretic sequence feature extraction, and k-mers

“…Since miRNA genesis depends on a pathway involving several protein complexes, structural features of pre-miRNAs have been found to be important [20]. Additionally, we have recently established the use of sequence motifs as features enabling the detection of pre-miRNAs [21,22]. Many machine learning models for pre-miRNA detection have been established using a variety of learning algorithms and training schemes [23][24][25][26].…”

“…They have in common that pri-miRNAs are transcribed from the genome and that hairpins (pre-miRNAs) are excised from these transcripts. Each pre-miRNA can have multiple mature miRNAs (18)(19)(20)(21)(22)(23)(24) nucleotides in length) which are incorporated into a protein complex, responsible for modulating the translation efficiency of multiple targets. MicroRNAs have been shown to exist in a variety of species ranging from viruses [2] to plants [3].…”