Nm (2′-O-methylation) is one of the most abundant modifications of mRNAs and non-coding RNAs occurring when a methyl group (-CH3) is added to the 2´ hydroxyl (-OH) of the ribose moiety. This modification can appear on any nucleotide (base) regardless of the type of nitrogenous base, because each ribose sugar has a hydroxyl group and so 2′-O-methyl ribose can occur on any base. Nm modification has a great contribution in many biological processes such as the normal functioning of tRNA, the protection of mRNA against degradation by DXO, and the biogenesis and specificity of rRNA. Recently, the single-molecule sequencing techniques for long reads of RNA sequences data offered by Oxford Nanopore technologies have enabled the direct detection of RNA modifications on the molecule that is being sequenced, but to our knowledge there were only two research attempts that applied this technology to predict the stoichiometry of Nm-modified sites in RNA sequence of yeast cells and 2′-O-Me subtypes in Hek293 human cell line. To this end, in this paper, we extend this research direction by proposing a bio-computational framework, Nm-Nano for predicting the existence Nm sites in Nanopore direct RNA sequencing reads of human cell lines. Nm-Nano framework integrates two supervised machine learning (ML) models for predicting Nm sites in Nanopore direct RNA sequencing data, namely the Extreme Gradient Boosting (XGBoost) and Random Forest (RF) with k-mers embedding models. The XGBoost is trained with the features extracted from the modified and unmodified Nanopore signals and their corresponding K-mers resulting from the reported underlying RNA sequence obtained by base-calling, while RF model is trained with the same set of features used to train the XGBoost, in addition to a dense vector representation of RNA k-mers generated by word2vec technique. The results on two benchmark data sets generated from Nanopore RNA sequencing data of Hela and Hek293 human cell lines show a great performance of Nm-Nano. In integrated validation testing, Nm-Nano has been able to identify Nm sites with a high accuracy of 99% and 92% using XGBoost and RF with k-mers embedding models respectively by training each model on 50% of a combination of Hela and Hek293 benchmark datasets and testing it for identifying Nm sites on the remaining 50% of the same combination. Deploying Nm-Nano to predict Nm sites in Hela cell line revealed that a total of 125 genes were identified as the top frequently Nm-modified genes among all other genes that have been modified by Nm sites in this cell line. The functional and gene set enrichment analysis on these identified genes in Hela cell line shows several high confidences (adjusted p-val < 0.05) enriched ontologies that were more representative of Nm modification role in immune response and cellular processes like: "C3HC4-type 370 RING finger domain binding", "Antigen processing and presentation (class I MHC)", and 371 "cytoplasmic translational initiation". Similarly, deploying Nm-Nano to predict Nm sites in Hek293 cell line revealed that a total of 61 genes were identified as the top frequently Nm-modified genes in this cell line. The functional and gene set enrichment analysis on these identified genes shows a wide range of functional processes like: "Glycolysis/Gluconeogenesis", "Regulation of protein localization to cell sur- 364 face", and "Aggrephagy" being significantly enriched that highlights the diverse regulatory role of Nm modifications, from their involvement in metabolic pathways, protein degradation and localization. The source code of Nm-Nano can be freely accessed at https://github.com/Janga-Lab/Nm-Nano.
