DaehwanKimLab / centrifuge

Classifier for metagenomic sequences
GNU General Public License v3.0
246 stars 73 forks source link

Low assigned reads #157

Open juulluu21 opened 5 years ago

juulluu21 commented 5 years ago

I have couple of pair ends reads. Average size of the libraries is 70 million (150nt paired end reads). However, only ~8% of the reads can be assigned to bacteria and virus (I used that index). Most of the reads remain as unassigned. Is it normal? Lots of the reads indeed are assigned to the host, but I am curious what % should map to bacteria/virus?

Thanks.

mourisl commented 5 years ago

That quite depends on the library preparation. If the cells are extracted from human, it is normal to see majority of the reads are from the host human.