Fig. 6
B. mori cocoon shell clustering validation. (a) PLS-DA analysis of the identified metabolites using cocoon shell age as independent variable40. The labels next to each circle, for this and all subsequent figures, indicates the samples’ names19. (b) PLS-DA analysis of the identified metabolites using cocoon shell colour (white/coloured) as independent variable40. (c) PLS-DA analysis of the identified metabolites using cocoon shell origin (Asian/European) as independent variable40. (d) PLS-DA analysis of the identified lipids using cocoon shell age as independent variable40. (e) PLS-DA analysis of the identified lipids using cocoon shell colour (white/coloured) as independent variable40. (f) PLS-DA analysis of the identified lipids using cocoon shell origin (Asian/European) as independent variable40. (g) Two-dimensional presentation of the results from the analysis of the normalized signal intensity values of the identified proteins using the Locally Linear Embedding (LLE) algorithm49. A 3D interactive map of the same results can be viewed in50. (h) Two-dimensional presentation of the results from the analysis of the normalized signal intensity values of the identified metabolites using the Locally Linear Embedding (LLE) algorithm49. A 3D interactive map of the same results can be viewed in50. (i) Two-dimensional presentation of the results from the analysis of the normalized signal intensity values of the identified lipids using the Locally Linear Embedding (LLE) algorithm49. A 3D interactive map of the same results can be viewed in50. (j) Hierarchical clustering dendrogram of the cocoon shell samples upon analysis of the normalized signal intensity values of the identified proteins (see Fig. 3a and67). The dotted circle indicates the two samples of cocoon shells of Japanese origin (one old cocoon shell sample and one of a Japanese hybrid strain). The asterisks (*) indicate samples with unusual clustering, i.e. old cocoon shells samples clustering with contemporary ones or vice versa (see19 for detains). (k) Hierarchical clustering dendrogram of the cocoon shell samples upon analysis of the normalized signal intensity values of the identified metabolites (see Fig. 4a and19,67 for detains). (l) Hierarchical clustering dendrogram of the cocoon shell samples upon analysis of the normalized signal intensity values of the identified lipids46,67. (m) Venn diagrams showing the grouping of the 80 cocoon shell samples that were analysed by proteomic, metabolomic and lipidomic analysis upon clustering of the samples using the k-means clustering algorithm. The clustering of the samples when k = 2 is shown68. The left and right Venn diagram differ by the clustering of the proteomics dataset by the k-means algorithm. The two cocoon shell samples on the left image are those samples closest to the centroid point of the clusters and the cocoon shell sample on the right image is the sample closest to the centroid point of the other clusters. (n) The World atlas depicted in Fig. 1a but this time layered with shaded areas and abstract lines that depict the expansion of silkworm races from Asia to Europe and their further reciprocal human-driven migration within Europe based on validation from historical manuscripts. The three indigo dots show the location of the three samples identified in the k-means clustering analyses shown in Fig. 6m.
