Remove microbes(cobionts) contamination: before or after insect de novo genome assembly?

Question

[hankool]

I'm going to work on a de novo insect genome assembly project. Since it's not practical to obtain a completely "clean" insect sample free of microbes (cobionts), bacterial and fungal reads may be generated alongside the insect genome sequencing reads. I'm unsure whether I should remove potential microbial contamination reads before assembly or after assembly.

If I remove them before assembly, possible lateral/horizontal gene transfers (LGT/HGT) might be mistakenly labeled as "contaminants" and removed erroneously without the genome context. On the other hand, if I remove them after assembly, I'm not sure whether these bacterial/fungal reads might be incorrectly assembled into normal insect contigs.

Answer 1

[Philipp Bayer]

I'd do it after assembly, just because computationally it's easier, and it's also less work to go through the decontamination results.

FCS-GX lets you specify which taxonomic units your assembly comes from and will then produce a report detailing which contig got removed why: https://github.com/ncbi/fcs/wiki/FCS-GX-quickstart

If you suspect HGT then you'll need to go through that report with a fine comb. Then again, most of the time what looks like HGT is just contamination, see https://genomebiology.biomedcentral.com/articles/10.1186/s13059-017-1214-2 and https://www.pnas.org/doi/full/10.1073/pnas.1525116113