Fig. 7

Metastasis-specific transcripts predict the metastasis and tissue origin of liver metastases. a. Barplot illustrating the statistical significance of 26 CRLM-SRTs. Metastasis-specific transcripts were defined as transcripts with a fold change greater than 10 between metastatic and nonmetastatic tumours. b. Heatmap displaying the expression of the indicated novel transcripts in tissues from individual CRLM patients, including data from Iso-seq and public dataset, as well as TCGA-CRC patients and normal colon and liver tissues. c. Area under the curve (AUC) estimation for the CRLM-SRT panel in the training and validation datasets, assessing its ability to discriminate CRLM from CRC. d. Boxplot demonstrating the number of detected SRTs by Iso-Seq in CRLM patients compared to LM patients with other tissue origins. e. AUC estimation for the CRLM-SRT panel in the training and validation sets, evaluating its effectiveness in identifying the tumour tissue origin of the colon. f. Barplot illustrating the statistical significance of 10 BCLM-SRTs. g. Heatmap displaying the expression of the indicated novel transcripts in tissues from individual BCLM patients, including data from Iso-seq and public datasets, as well as TCGA-BRCA patients and normal breast and liver tissues. h. AUC estimation for the BCLM-SRT panel in the training and validation sets, assessing its ability to discriminate BCLM from breast cancer. i. Boxplot illustrating the number of detected SRTs by Iso-Seq in BCLM patients compared to LM patients with other tissue origins. j. AUC estimation for the BCLM-SRT panel in the training and validation sets, evaluating its effectiveness in identifying the tumour tissue origin of the breast