Hi Yang. Random integration using CRISPR in mice seems to be rare. GeneCopoeia provides CRISPR transgenic mouse services, and in our hands 1) The frequency of correct donor insertion is about 3-5%, and 2) We almost never see random integration. This makes sense. If CRISPR is working, the frequency of integration will be stimulated by several orders of magnitude, so if you are screening about 100-200 mice, then the ffrequency of random integration will be below the levele of detection. This is opposed to cell lines, in which a selectable marker is placed on a donor plasmid. In that case, you are selecting for cells that are drug resistant, and the only way they can survive is to integrate the plasmid, whether randomly or correctly. That is why people can still detect drug resistant cells in control transfections of donor but no CRISPR. If you want to know more, visit us at http://www.genecopoeia.com/resource/technical-and-application-notes/, where we have technical literature on genome editing topics.