Keith Hornberger’s Post

A new edition of Ramblings is up on the blog. This month I cover the power of GRCs, how LinkedIn is (mis)used, the intrusion of proper nouns into science, a bit more on ML model training, and more. Check it out! 📚

A new Ramblings post is up on the blog. This month I cover: some of the downsides of industry life, time management, the value of saying thanks, the value of knowing what you don’t know, and more - check it out! 📚

For your Sunday coffee ☕️ a new edition of Ramblings is up on the blog. This month I cover: making a good slide deck, training sets for AI/ML models, the burden of clinical trial failure rates, the importance of good pharmacology and doing control experiments in the literature, and more. Check it out!

Wow - amazing! Congrats to my coauthor Erika Vieira Araujo!

This is an extremely welcome change! Biological data is log-normally distributed and therefore best described by geometric means and standard deviations. Doing so will prevent the emergence of nonsensical negative values in the error range. An easier way to express this, though, is as proposed below — log transform the data as pXC50 (-log10 XC50) and then you can use arithmetic means and standard deviations again. It also prevents your brain from falling into the trap of thinking that an IC50 of 10 nM and 30 nM are different from each other, when the error in most biological assays is ~half a log. You can make really bad SAR decisions by trying to squint at (logarithmically) small changes and convince yourself they’re different from each other. I wrote a whole thread about this topic on the other app, link below. https://lnkd.in/e-QBhDd7

We are hiring a senior medicinal chemistry project leader in our neuroscience group - check it out and consider applying!

Just a quick note to say that I’m overwhelmed by the well wishes from my network about this post — literally! I’ve received so many DMs that I can’t respond to all of them, as much as I’d like to. Please know that I saw your messages and appreciate all the support. 🙏

Congrats on an amazing run in the big chair and enjoy your next chapter Larry! Appreciate your mentorship and leadership over the years. You took a small nucleus and built a world-leading #TPD medicinal chemistry department that has consistently delivered clinical candidates. I have some big shoes to fill.