As an illustration, we apply LR, RF and TRF to the C-to-U conversion data previously investigated in relation to random forest in the bioinformatics literature [14, 40]. In summary, RNA editing is the process whereby RNA is modified from the sequence of the corresponding DNA template [40]. For instance, cytidine-to-uridine conversion (abbreviated C-to-U conversion) is common in plant mitochondria. Cummings and Myers [40] suggest to use information from neighboring sequence regions flanking the sites of interest to predict editing status, among others in Arabidopsis thaliana. For each of the 876 complete observations included in the dataset (available at https://static-content.springer.com/esm/art%3A10.1186%2F1471-2105-5-132/MediaObjects/12859_2004_248_MOESM1_ESM.txt), the following features are available: the binary response at the site of interest (edited versus not edited)the 40 nucleotides at positions -20 to 20, relative to the edited site (4 categories: A, C, T, G), whereby we consider only the nucleotides at positions -5 to 5 as candidates in the present study,the codon position cp (4 categories: P0, P1, P2, PX),the (continuous) estimated folding energy (fe)the (continuous) difference dfe in estimated folding energy between pre-edited and edited sequences.
