Fig. 6

Multiomics integration of data from different papers based on the same animal model using sparse Projection of Latent Structures – Discriminant Analysis (sPLS-DA). Individual block plots of samples in the two-dimensional space spawned by the 2 first Latent Structures of a Generation F0, b Generation F1, and c Generation F2. sPLS-DA requires an equal number of samples in each block (dataset), thus the most representative samples according to metabolomic and lipidomic profile were selected using 2-omics PLS. The closest samples to the centroid of the group were selected to match the number of sncRNA-sequenced samples. Due to the integration of more datasets, the clustering of samples belonging to the same group (CTRL, HFD, and HFDt) is enhanced, particularly in the F1 generation. The block correlation plots of d Generation F0, e Generation F1, and f Generation F2 highlight each datasets share most similarities among each other, based on the first latent structure. Low correlation values suggest sources of differences among samples from different groups. Number of replicates: F0_CTRL = 3, F0_HFD = 3, F0_HFDt = 3; F1_CTRL = 2, F1_HFD = 2, F1_HFDt = 2; F2_CTRL = 3, F2_HFD = 3, F2_HFDt = 3