I am posting this question for my friend.
He is analyzing his CHIP DNA solid deep sequence data, and find out that near 80% reads can not be mapped to the human genome. We are wondering if this high percentage unmapped reads is normal in CHIP DNA deep sequence or there may be something wrong with his result.
Your 20% mapping yield looks like low for normal ChIP experiment, even for human. Several factors can reduce this mapping yield. I am wondering which kind of ChIP was used in your case. That is, which kind of proteins was ChIPed?
I recall that our first samples that we ran on the Solid sequencer have had bad performance. Not quite an 80% loss but around 40%-60% reads were unmappable (yeast). Some other lab members will hopefully chime in with more details.