WACS: improving ChIP-seq peak calling by optimally weighting controls.

摘要: Chromatin immunoprecipitation followed by high throughput sequencing (ChIP-seq), initially introduced more than a decade ago, is widely used the scientific community to detect protein/DNA binding and histone modifications across genome. Every experiment prone noise bias, ChIP-seq experiments are no exception. To alleviate incorporation of control datasets in analysis an essential step. The controls account for background signal, while remainder signal captures true or modification. However, recurrent issue different types bias experiments. Depending on which used, aspects better worse accounted for, peak calling can produce results same experiment. Consequently, generating “smart” controls, model non-signal effect specific experiment, could enhance contrast increase reliability reproducibility results. We propose algorithm, Weighted Analysis (WACS), extension well-known caller MACS2. There two main steps WACS: First, weights estimated each using non-negative least squares regression. goal customize distribution This then calling. demonstrate that WACS significantly outperforms MACS2 AIControl, another recent algorithm smart detection enriched regions along genome, terms motif enrichment analyses. ultimately improves our understanding their biases, shows approximation controls.