Low-quality datasets for QC purposes?

Question

[Bohdan Khomtchouk]

Has anyone ever worked with any cancer ChIP-seq datasets (e.g., from TCGA) that don't replicate well or are of not-so-great quality? I'm testing out QC and want to show what a bad sample looks like. Cancer is preferred but not necessary, especially if anyone has seen cases that have jumped out as being particularly low quality and/or difficult to replicate.

Answer 1

[dyollluap]

Try some of the ENCODE ChIP-seq datasets that have inconsistent results:

https://www.encodeproject.org/search/?audit.ERROR.category=inconsistent+control

The project has good documentation, annotation and pre-release validation standards, so identifies datasets that are sub-optimal.