Table 7.

MiPepid’s prediction on the non-high-confidence data in SmProt

Data source	#sORFs	avg sORF length (aa)	#Predicted positive	Proportion
high-throughput literature mining	25,663	44	20,516	0.80
ribosome profiling	13,715	36	8596	0.63
MS data	324	15	233	0.72

high-throughput literature mining: published sORFs that were identified using high-throughput experimental methods;

ribosome profiling: sORFs predicted from Ribo-Seq data;

MS data: sORFs predicted from MS data;

#sORFs: number of sORFs from a particular data source;

avg sORF length (aa): the average length of sORFs measured in number of amino acids;

#predicted positive: number of sORFs that are predicted as positive by MiPepid;

proportion: $\frac{\mathrm{avg}\text{sORF length}}{\#\text{predicted positive}}$