16s analysis MySeq PE - analysis of only forward reads?

Question

[manekineko]

Dear all, I have information from a company that 16s V3-V5 metagenomics data from MiSeq PE 2x300 and they analyze only the Forward reads from the sample (no joining the reads). I was wondering which approach can be more accurate - joining the reads and losing some reads or using only the forwards?

Answer 1

[Asaf]

If you loose reads when joining paired reads it means that they were problematic to start with, you only disregarded this information. I think you'll get more reliable and accurate results using the paired-end data