average number of duplicated reads for NGS

Question

[igorm]

Hi,

I am looking for some wisdom ngs regarding read duplication.

What is the expected/acceptable number of duplicated reads when sequencing human dna, for example at average coverage 1x? Are there any official statistic on this topic? How to improve the protocol to minimize the duplicated reads?

Thanks.

[GenoMax]

expected/acceptable number of duplicated reads when sequencing human dna, for example at average coverage 1x

Don't think there is an expected/acceptable number. One would always want to minimize PCR/technical duplicates. Some of this is going to depend on quality of the libraries. There are going to applications where duplication would be normal/expected (RNAseq etc).

How to improve the protocol to minimize the duplicated reads?

There are PCR-free methods/kits that minimize duplicates. There will likely be optical/clustering duplicates depending on type of sequencer used (Duplicates on Illumina ).

[igorm]

Thanks for fast reply GenoMax and advice. Based on your experience what is the percentage that would clearly seem unexpected? For example 10% duplicates at 1x coverage on PCR library reasonable or not?