Entering edit mode
3.2 years ago
Sus
▴
40
Hello,
I have a few samples sequenced using in-house primers and I often have more than 60% of the reads being contamination. For instance the last samples we received were around 70% contamination. It does not seem to bother some of my coworkers more than that since they're always satisfied with just analyzing the 30% that worked.
It seems a bit odd to me to analyze data were most of it is not even the expected organism (and not even a close parent). In my head if 80% of the reads are contaminated then there's a high chance that the remaining 20% are not really trustworthy. Does this make sense and if yes then how much contamination is too much ?
How are you gauging what is and is not contamination? - BLAST of a representative sub-sample of reads?
The more important question first of all imho is what this all is about: What did you sequence, what is the setup? This information is completely lacking. Please give context.