Summary: ChIA-PET is rapidly emerging as an important experimental approach to detect chromatin long-range interactions at high resolution. Here, we present Model based Interaction Calling from ChIA-PET data (MICC), an easy-to-use R package to detect chromatin interactions from ChIA-PET sequencing data. By applying a Bayesian mixture model to systematically remove random ligation and random collision noise, MICC could identify chromatin interactions with a significantly higher sensitivity than existing methods at the same false discovery rate.Availability and implementation: http://bioinfo.au.tsinghua.edu.cn/member/xwwang/MICCusageContact: michael.zhang@utdallas.edu or xwwang@tsinghua.edu.cnSupplementary information: Supplementary data are available at Bioinformatics online.