Fig. 5

Biogeographical abundance profiles of MAGs. Heatmap plots showing the abundance of recruited reads from various samples and sample groups to each of the 30 MAG clusters as well as to a co-assembly of all samples in the time-series (“BalticAsm”) at the (a) 85 % and (b) 99 % identity cutoff levels. Shown values represent number of recruited reads/kb of genome per 10,000 queried reads. For clarity, several sample groups have been collapsed with recruitment values averaged over samples in the group. Sample groups are indicated by the lower color strip above the plot and samples are ordered by salinity (shown in the upper color strip). See Fig. S11 in Additional file 1 for full visualizations of samples