figshare
Grevesse, Thomas;
Guéguen, Céline;
Onana, Vera E.;
Walsh, David A.
—
2023
Additional file 1: Figure S1. Estimation of the rank for the NMF analysis of the EC abundance matrices annotated from metagenomes (top panels) and metatranscriptomes (bottom panels). Left panels: evolution of various parameters as a function of the rank used in the NMF analysis. Cophenetic correlation represents the correlation between the sample distances from the consensus matrix and the cophenetic distance between these samples when they are clustered. The rss is the residual sum of squares between the original EC abundance matrix and its estimate using the NMF algorithm. The dispersion is defined as 1-rss/Σi,j (Vi,j)2 (Vi,j are the entries of the EC abundance matrix) and estimates the fraction of variance of the EC abundance matrix explained by the NMF results. Residuals is the sum of residuals between the original EC abundance matrix and the matrix estimated using the NMF. Right panels: consensus matrices based on clustering the coefficient matrices at each of the 100 runs of the NMF analysis. The heatmap represents the fraction of times 2 samples fall in the same clusters out of 100 runs. Figure S2. Heatmaps of the basis matrix (left) and coefficient matrix (right) obtained after running an NMF analysis on the EC abundance matrix annotated from metagenomes and using a rank value of 4. Figure S3. Heatmaps of the basis matrix (left) and coefficient matrix (right) obtained after running an NMF analysis on the EC abundance matrix annotated from metatranscriptomes and using a rank value of 4. Figure S4. Lignin-derived aromatic compound degradation pathways completeness in the 4 water column features (surface, SCM, FDOMmax, and deep water) for metagenomes (left) and metatranscriptomes (right). Figure S5. Normalized abundance per water column feature of KO number markers of aromatic compounds degradation pathways annotated from metagenomes. Figure S6. Normalized abundance per water column feature of KO numbers markers of aromatic compounds degradation pathways annotated from metatranscriptomes. Figure S7. Estimated fraction of the microbiome harboring genes annotated with KO marker of aromatic compounds. Figure S8. Completeness and contamination of the set of 1772 MAGs reconstructed from 22 individual metagenomes. The vertical and horizontal dotted lines represent 10% contamination and 30% completeness respectively. Selected MAGs are the 46 MAGs that have been selected as most implicated in lignin-derived aromatic compound degradation based on the number and completeness of their aromatic compound degradation pathways. Figure S9. Taxonomic identity of the MAGs harboring genes annotated with KO marker of aromatic compounds degradation pathways. The taxonomy is displayed at the class level. Figure S10. Phylogenetic tree reconstructed from the concatenation of 120 conserved genes for the bacterial genomes of the MAGs dataset. The heatmap represents the completeness of the lignin-derived aromatic compounds degradation funneling pathways. Red stars correspond to the MAGs selected based on the amount and completeness of pathways they harbor. Figure S11. Taxonomic identity of the MAGs most implicated in the degradation of lignin-derived aromatic compounds. Taxonomy is displayed at the order level and based on the tree placement of the MAGs, using the GTDB. Figure S12. Heatmap of the average nucleotide identity for the 46 MAGs most implicated in the degradation of lignin-derived aromatic compounds.
https://creativecommons.org/licenses/by/4.0/legalcode