Closed abremges closed 6 years ago
I stumbled upon this looking at boxplot_completeness
, where (I believe) each bins has equal weight and thus—in my case—the depicted median etc. was zero.
Implemented in bce66ecac88490be3f5d6102af226db5383bdc99. The new option is --min_length, -n (default 0 to retain default behavior).
Great, thank you! Will test this new feature soon-ish. 👍
First of all, great work with AMBER!
I ran AMBER on the mouse gut toy dataset, which contains many very small contigs in the GSA. Some bins exclusively contain small contigs and are not recoverable by common genome binners.
It is already possible to manually exclude a set of genomes; I propose a complementary feature: Filter contigs by size (threshold set by user, default e.g. 2.5kb) and exclude these from analyses. This will also remove some gold standard bins completely (if they don't contain longer contigs).