$BEAM 60 mg: true mean M-AAT reached 94% of total AAT, compared with 80% for the MZ genotype.
$WVE 400 mg MAD: mean max M-AAT reached 58.7% of total AAT, compared with 64% for the MZ genotype.
One set a high bar and beat it.
One set a low bar and still could not beat it—even after inflating the data. 🤣
$WVE the more I'm looking at the AATD data, the more alarmed I grow.
It looks to me as if total serum AAT stays the same, if not is lowered following wve-006 treatment. And it now makes sense why Co does not show total serum AAT for entire cohort (they show Z-AAT reduction instead as well as isolated M-AAT levels).
$beam $krro
If you invest in $WVE, you do it because of INHBE (and platform optionality), but NOT AATD*.
*not investment advice
搞医学研究的朋友应该深有体会,从文献检索、方案设计到数据分析、论文写作。
每个环节都想借助 AI 提效,但普通 AI 缺乏严谨的医学逻辑,用起来总觉得不太靠谱。
最近看到 AIPOCH Medical Research Skills 这个项目,提供了 500 多个专为医学研究设计的 AI 智能体技能库。
它把医学研究拆成证据洞察、方案设计、数据分析、学术写作四大模块。
每个技能都内嵌了专业的医学研究逻辑,比如文献真实性约束、研究类型识别等。
GitHub:https://t.co/aA32tMXjDG
技能之间可以自由组合,从单任务执行到多步骤流水线都能搭建。
还提供一套质量评估框架,��在技能上线前进行全面的质量审核,确保输出的可靠性。
如果你正在用 AI 辅助医学研究,这个技能库值得试试,目前��在持续更新中。