Teytelman, L. et al. Proc. Natl. Acad. Sci. USA 110, 18602–18607 (2013).

Even well-established methods, such as chromatin immunoprecipitation followed by deep sequencing (ChIP-seq), are not immune to inherent bias, as Teytelman et al. discovered when using ChIP-seq to look at where in the yeast genome a protein complex that represses transcription binds. To their surprise, the researchers found components of the complex associated with a lot of highly expressed genes, a finding not in line with the known function of this complex. Following up with analyses of other proteins, Teytelman et al. discovered that highly expressed loci were always enriched in ChIP peaks, regardless of which protein was pulled down. They caution the scientific community that any ChIP-seq result should be checked against the expression of the locus and that unexpected findings need to be validated independently.