hclust2.py heatmap in MetaPhlAn3 --ftop features

Question

[serene.s]

Hi Guys! If anyone of you has used metaphlan3 for the shotgun metagenomic datasets and made the heatmaps using their inbuilt script called hclust2.py then I have a question. There is a flag ---ftop n (n is the integer which you want to select as top features/taxa), I want to know how do they select the top n features based on the percentile rank of the relative abundance values of the particular taxa/features among all the samples. Please explain if anyone knows it.

Thanks Saraswati Awasthi