Here's my take on the protein sequences encoded by the alleles reported at #geneeditsummit for #CRISPRbabies. Lulu's mutation (delta15) leads to a five amino acid deletion, but the protein largely resembles wild type. The consequence of this mutation is unknown... (1) Thread 
Nana has two different mutations (excluding possible mosaicity). The first (ins1) leads to a variant that is similar to the well-studied delta 32 variant, except it has an additional 11 amino acids that are neither in WT or delta 32... (2) Thread
Her second mutation (delta 4) produces an even shorter protein than wild-type or delta 32. It also contains nine amino acids not present in any of the other variants. The phenotypic consequence of either of these mutations (ins1 or delta4) is unknown to my knowledge. (3) Thread
This is what really bothers me. The children are test subjects for protein variants that haven't been vetted in animals. Any of the variants could have unintended consequences. Dominant negative, gain of function, neomorphic, etc... (4) Thread
So in addition to the dubious medical rationale for these experiments, there was nowhere near enough supporting data from animal studies to show that the mutations would be helpful rather than harmful. Just hunches, hopes, and a cavalier attitude... (5) Thread
This lazy approach towards genome modification in humans is frankly frightening. I question the motivations of Prof. He. I question his good judgement. I question the oversight system that enabled these experiments to happen undetected... (6) Thread
I hope that future efforts -- if any -- will be transparent and highly scrutinized by an international panel of experts and ethicists PRIOR to pregnancy. The technology holds great promise. But the risks are all too real as well. (7) / Thread
UPDATE: I used an outdated model of the membrane topology to make my original figure. Here is an corrected version. Apologies for the error (8) / Thread.
