search

RasmussenLab / vamb

Variational autoencoder for metagenomic binning
MIT License
241 stars 44 forks source link

eukaryotic MAGs #115

Open Thexiyang opened 2 years ago

Thexiyang commented 2 years ago

Thanks for this nice too! Does this tool also work for binning eukaryotic MAGs?

jakobnissen commented 2 years ago

We haven't tested it yet for eukaryotes. Presumably yes, it should work, because eukaryote MAGs also follow a certain co-abundance and kmer composition. But the devil may lie in the details, and there may be important nuances that make eukaryotic MAGs much harder to bin.

We could think about adding a benchmark to a eukaryote-heavy dataset, if we can find a good synthetic one. I'll keep the issue open.

Andyargueasae commented 4 months ago

Hello Jakob,

I am also interested in binning euMAGs, but here I found a problem with the workflow: when my input data contains only one row in samples2data.tsv file (i.e.: I only got one sample in the dataset), the workflow reports error pointing to line 220, and says the index is out of range (while in that rule the only indexing command is the input): image And here is the error: image Does that mean that avamb doesn't support single-sample binning?

Cheers