What causes very low read mapping? CUT&RUN with Bowtie2

Question

[smurph50]

Generally, what causes low read mapping efficiency? Some samples have 70% and others are 8-20%. I am experienced with bowtie2 and tested with/without trimming. These are human and mouse cut&run.

We are not confident in the antibodies used, but I thought bad antibodies would lead to no peaks later on rather than low alignment.

[GenoMax]

No magic bullet here. Get representative reads that do not map and investigate with BLAST. You may have some unexpected contamination.

Answer 1

[ATpoint]

Much more likely that IP failed and the library is simply a mess full of adapter dimers and ligation/PCR artifacts.

[GenoMax]

tested with/without trimming

Assuming trimming was done properly so perhaps the latter based on your experience?

[ATpoint]

pretty much, yes. contamination in the lab in terms of fungi or bacteria are rare in my experience. blast will just by chance give some random hits to any bacteria but chance these really were in your tube are almost zero. with these chip-like libraries and poor antibodies a failed ip is almost always the best explanation.