. 2024 May 23;7(8):e202402730. doi: 10.26508/lsa.202402730

Table 1.

Brief overview of the datasets used in the study.

Dataset	Public ID	Sample preservation	n	Stage	FFPE-RNA isolation protocol	FFPE library preparation protocol	Gene-expression platform
Discovery cohort – low-quality set	GSE33113	FFPE and matched FF	22	II	RNeasy FFPE kit (QIAGEN)	NEBNext Ultra Directional RNA Library Prep Kit with rRNA depletion (Illumina)	FFPE: HiSeq4000, 150 bp, paired-end (Illumina)
Discovery cohort – low-quality set	GSE33113	FFPE and matched FF	22	II	RNeasy FFPE kit (QIAGEN)		FF: Affymetrix U133 Plus 2.0 chips
Discovery cohort – moderate-quality set	GSE33113	FFPE and matched FF	12	II	Allprep DNA/RNA/miRNA Universal kit (QIAGEN)	Truseq RNA Exome Prep kit (Illumina)	FFPE: HiSeq2500, 65 bp, single-end (Illumina)
Discovery cohort – moderate-quality set	GSE33113	FFPE and matched FF	12	II	RNeasy FFPE kit (QIAGEN)	Truseq RNA Exome Prep kit (Illumina)	FF: Affymetrix U133 Plus 2.0 chips
Training cohort	TCGA COAD and READ	FF	622	I–IV	NA	NA	Multiple
FF-RNA validation cohort	GSE39582 (CIT cohort)	FF	519	I–IV	NA	NA	Affymetrix U133 Plus 2.0 chips
FFPE-RNA application cohort	GSE267010	FFPE	104	IV	Allprep DNA/RNA FFPE kit (QIAGEN)	SMARTer Stranded Total RNA-Seq Kit v3 Pico (Takara Bio [TAKARA])	NovaSeq6000, 150 bp, paired-end (Illumina)

ID, identification; n, number of patients; FFPE, formalin-fixed paraffin-embedded; FF, fresh-frozen; TCGA, The Cancer Genome Atlas; COAD, colon adenocarcinomas; READ, rectum adenocarcinomas; NA, not applicable; CIT, Cartes d’Identité des Tumeurs.