This is common with microbial datasets and the best solution is to apply a threshold, to only show those that are at least x% in abundance or occur in x number of samples, or both of these.
filtered = genefilter_sample(physeq_herschel_rel, filterfun_sample(function(x) x > 0.01), A=0.05*nsamples(physeq_herschel_rel))
This looks much better, but it is hard to distinguish between some of the colours. This is an inherent problem with barplots but we can try our best to make it a bit clearer.
## Hard to distinguish these colours, so let's try changing the colour scheme
## First we extract the data from the plot above
## Then we will assign colours to families and replot
## This time for the plotting we will use ggplot itself (which is what phyloseq was using above)
##First we extract the data from the plot above
## Then we will assign colours to families and replot
## This time for the plotting we will use ggplot itself (which is what phyloseq was using above)