Table 1 Various terms used in assessing the effect of freeze-thaw on RNA-sequencing, their definitions, and the specific analyses they are applied to

RNA sequencing

Distortion

A generic term referring to changes in RNA-sequencing data introduced due to technical factors.

Consistency

A generic term referring to the reproducibility of RNA-sequencing results between samples.

Sample Quality

Noise (randomness)

The fraction of reads in a sample that are randomly counted, rather than mapping to a sample-specific gene.

Results section 2 and 3

Freeze-thaw

The number of freeze-thaw cycles a sample undergoes. A freeze-thaw cycle is defined as freezing a sample in − 80 °C for at least 24 h, proceeded by thawing it to room temperature, with the first hour spent on ice.

All results sections

RIN

The RNA integrity number as previously decsibed [23]

All results sections

DE Reproducibility

Similarity

Spearman correlation of LFC results from differential expression on sample subsets. Correlation was taken between all pairs of subsets.

Results section 4 and 5

Discordance

Standard deviation of LFC results from differential expression on sample subsets. Standard deviation was taken across all subsets for each gene.

Results section 4 and 5

Bias

3′ Bias

The extent to which reads map in a skewed manner to the 3′ end of a transcript.

Results section 1 and 6

Median coverage percentile

The nucleotide percentile (relative to transcript length) at which median cumulative coverage across a transcript is achieved; cumulative coverage is aggregated from the 5′ end to the 3′ end. This is a measure of bias in which a larger median coverage percentile indicates more 3′ bias and vice versa

Results section 1 and 6