One of the best critiques of the Nature Methods article was in the Wired piece, pulling it out here for quick reference: http://blogs.sciencemag.org/pipeline/a

One of the best critiques of the Nature Methods article was in the Wired piece, pulling it out here for quick reference: http://blogs.sciencemag.org/pipeline/archives/2017/05/31/trouble-with-crispr-maybe-but-maybe-not http://blogs.sciencemag.org/pipeline/archives/2017/05/31/tro... Derek Lowe is an awesome writer about these kinds of issues. He highlights some of the best of criticisms out there, notably that it's not clear whether the increased rate of apparent mutation and the striking amount of mutations shared between the two treated animals was due to the "experimental" animals being siblings, but the "control" animal being more distantly related. This article isn't the final word by far, and I think the authors didn't claim it to be either. They wanted to put a concern out there for the scientific community to investigate, as well they should, and the community will figure this out. Remember though, CRISPR is by definition a mutagen. It does create rather substantial disruption in DNA structure (double stranded breaks), and relies on intrinsic DNA repair processes to affect the desired end result. This is likely to be a somewhat noisy process, and we shouldn't be terribly surprised if unexpected changes do occur. Indeed the scientific community is well aware of these limitations, and I think the more popular press shouldn't sell CRISPR as equivalent to programming DNA in the same way we write computer programs.