Table 2. The parameters of the public tools used in RNAMiner, the parameter values, and the descriptions.

Tool	Parameter	Value	Description
TopHat2	—read-mismatches	2	The maximum number of mismatched nucleotides between a read and a reference allowed for a valid mapping.
	—read-gap-length	2	The maximum number of gaps in the alignment between a read and a reference genome allowed for a valid mapping.
	—splice-mismatches	0	The maximum number of mismatches allowed in the "anchor" region of a spliced alignment.
	—segment-mismatches	2	Read segments are mapped independently, allowing up to this number of mismatches in each segment alignment.
	—segment-length	25	A read is cut into segments each having at least this length. These segments are mapped independently.
Bowtie2	—end-to-end		In this mode, Bowtie2 requires that the entire read to be aligned from one end to the other, without any trimming (or "soft clipping") of characters from either end. Local alignment is not used in Bowtie2.
	—sensitive		This option generally balances speed, sensitivity and accuracy.
Cufflinks	—frag-len-mean	200	The expected (mean) fragment length.
	—frag-len-std-dev	80	The standard deviation of fragment lengths.
	—min-isoform-fraction	0.10	Suppress isoform transcripts below this abundance level.
	—pre-mrna-fraction	0.15	Suppress intra-intronic transcripts below this level.
	—max-intron-length	300000	Ignore alignments with gaps longer than this.
Cuffdiff	—min-alignment-count	10	Minimum number of alignments in a locus for testing.
	—FDR	0.05	The maximum false discovery rate allowed after statistical correction.
	—frag-len-mean	200	The expected (mean) fragment length.
	—frag-len-std-dev	80	The standard deviation of fragment lengths.
HTSeq	-a	10	Skip all reads with alignment quality lower than this minimum value.
	-t	Exon	Feature type (3rd column in GFF file) to be used. All the features of other types are ignored.
	-i	gene_id	GFF attribute to be used as feature ID.
	-m	Union	Mode to handle reads overlapping more than one feature.
DESeq2	Test	LRT	Use the likelihood ratio test on the difference in deviation between a full and reduced model formula (nbinomLRT).
	fitType	parametric	The type of fitting of dispersions to the mean intensity. Parametric: fit a dispersion-mean relation via a robust gamma-family GLM.
edgeR	Pair	NULL	First two levels of object$samples$group (a factor) are used in comparison.
	Dispersion	NULL	Either the common or tagwise dispersion estimates from the DGEList object will be used, according to the value of common.disp.
	common.disp	TRUE	Testing carried out by common dispersion for each tag/gene.