Table 4:

Metrics for Challenge 2 evaluation

Metrics	Description
Irreproducibility Measure (IM) and Area under the Coefficient of Variation Curve (ACVC)	IM and ACVC characterize the coefficient of variation of abundance estimates among multiple replicates.
Consistency Measure (CM) and Area under the Consistency Curve (ACC)	CM and ACC characterize the similarity of abundance profiles between pairs of replicates.
Resolution Entropy (RE)	RE characterizes the resolution of abundance estimation.
Spearman Correlation Coefficient (SCC)	SCC evaluates the monotonic relationship between the estimation and the ground truth.
Median Relative Difference (MRD)	MRD is the median of the relative difference of abundance estimates among all transcripts.
Normalized Root Mean Square Error (NRMSE)	NRMSE measures the normalized root mean square error between the estimation and the ground truth, which characterizes the variability of the quantification accuracy.
Percentage of Expressed Transcripts (PET)	PET characterizes the percentage of truly expressed transcripts in SIRV-set4 data