Table 2. Benchmarking machine learning methods for coding potential prediction based on trinucleotides count.

F1-score for each one of the 15 species in which the algorithms were tested. Other metrics (sensitivity, specificity, precision, accuracy and the confusion matrix) used for the comparison of the algorithm’s performance were made available at the Extended data: Supplementary File S2 ¹⁹ .

Species	ANN	CNN	K-NN	NAIVE BAYES	RANDOM FOREST	SVM	XGBoost
Anolis carolinensis	98.47	98.31	93.55	95.50	98.30	98.03	98.79
Chrysemys picta bellii	96.54	96.02	93.54	93.13	96.89	96.04	98.00
Crocodylus porosus	96.74	96.48	93.67	93.93	97.26	96.35	98.15
Danio rerio	97.54	97.77	95.44	94.55	97.56	97.27	97.98
Eptatretus burgeri	94.88	95.69	92.24	94.57	97.35	95.82	97.56
Gallus gallus	98.47	98.27	96.87	95.11	98.91	98.06	99.24
Homo sapiens	98.01	97.66	96.63	86.00	98.30	96.83	98.50
Latimeria chalumnae	99.05	98.72	91.61	98.23	99.56	99.24	99.57
Monodelphis domestica	98.39	98.09	97.11	95.31	98.67	98.01	98.84
Mus musculus	96.67	96.96	95.95	91.56	97.66	96.10	97.73
Notechis scutatus	95.90	94.10	87.77	89,81	94.94	95.73	96.51
Ornithorhynchus anatinus	97.23	96.59	93.59	91.45	96.99	96.38	97.61
Petromyzon marinus	98.40	98.26	88.10	95.99	98.79	97.49	99.42
Sphenodon punctatus	97.83	96.97	78.41	96.70	96.46	95.29	99.20
Xenopus tropicalis	98.28	98.81	85.53	97.14	98.88	97.20	99.13