doi: 10.1016/j.gendis.2025.101537.
eCollection 2025 Nov.
A multi-omic integrative approach combining m6A-epitranscriptomic, transcriptomic, and splicing alternative events reveals potential candidates for colorectal cancer diagnosis
Affiliations
- PMID: 40808875
- PMCID: PMC12343466
- DOI: 10.1016/j.gendis.2025.101537
Item in Clipboard
A multi-omic integrative approach combining m6A-epitranscriptomic, transcriptomic, and splicing alternative events reveals potential candidates for colorectal cancer diagnosis
Genes Dis.
.
No abstract available
Conflict of interest statement
The authors declared no conflict of interests.
Figures
Multiomic integrative analysis combining m6A-epitranscriptomic, transcriptomic, and AS events in leukocytes in CRC. (A) PCA of m6A analysis plotting was conducted using transformed data on the log scale normalized to library size using the DESeq2 package. A variance stabilizing transformation was applied to remove the dependence of variance on the mean, particularly addressing the high variance of the log counts when the mean is low. The percentage of global variation explained by each principal component is provided in the axis labels. (B) Metagene profile of m6A distribution across the transcriptome in controls and patients with CRC. (C) Normalized counts of m6A immunoprecipitated RNA, divided by 5′ UTR, 3′ UTR, and CDS regions. Asterisks indicate significant differences between the groups according to the Wilcoxon test (∗p < 0.05, ∗∗p < 0.01, ∗∗∗p < 0.001, ∗∗∗∗p < 0.0001). (D) The distribution of the m6A peaks along the transcript, including the 5′ UTR, 3′ UTR, first and other exons, as well as first and other introns. We used the ChIPseeker R package to annotate the genomic region of the peak. (E) Gene Set Enrichment Analysis was done using the following genes: For m6A, we applied an FDR <0.001 with an absolute LogFC greater than 2.5, and genes that had more than three m6A peaks (479 genes). For mRNA, we identified genes with an absolute LogFC greater than or equal to 1.2 and a p-value ≤0.01 (450 genes). For AS, we applied an absolute LogFC greater than 2 and a p-value <0.001 (291 genes). (F) Arrow plot from multiblock sPLS-DA performed on the data of integrated dataset of epitranscriptomic, transcriptomic, and AS events. The samples are projected into the space spanned by the first two components for each dataset and then overlaid across datasets. The start (tail) of the arrow indicates the location of the centroid of all datasets (blocks) for samples. The end (tip) of each arrow indicates the location of samples in each block, projected onto the averaged latent components. (G) sPLS-DA consensus plot for the combination of the three datasets showing complete discrimination of groups of datasets. (H) Sample scatterplot from plotDiablo displaying Pearson correlation between each component (lower diagonal plot). (I) Random forest analysis of the most important variables for predicting the presence and the absence of CRC. (J) ROC curve analysis of m6A, Mrna, and AS analysis for the prediction of CRC. AS, alternative splicing; CRC, colorectal cancer; FDR, false discovery rate; LogFC, log fold change; NES, normalized enrichment score; PC, principal component; PCA, principal component analysis; ROC, reciever operating curve; sPLS-DA, sparse partial least squares regression for discrimination analysis; TSS, transcription start site; TTS, transcription termination site; UTR, untranslated region.
References
LinkOut - more resources
Full Text Sources
