Tweetorial: Rule of 5
30 July 2021
Hoo boy. This was a jumping-off point to talk a bit about PROTAC physchem property space, as PROTACs clearly violate the traditional “Rule of 5”. As a discipline, I’d like to believe we’ve moved on from a rigid following of this so-called “rule” — certainly in PROTACland, we had to do that a long time ago. Some good discussion got rolling on this one as some folks clearly get a little triggered by things like Ro5. Apparently there’s still swaths of chemists out there who cling to their beloved Ro5. Idk.
Also some crosslinks here to discussions with Bob the Grumpy Med Chemist, who has rapidly become one of my best buds on Twitter.
🧵Few drug discovery concepts are as misunderstood and wielded as a blunt instrument as the Rule of 5 (Ro5). The rule is not a rule, and better applied as a guideline. Thoughts on Ro5 and its relation to PROTACs follow. 1/
sciencedirect.com/science/articl…
Medicinal chemists delight in breaking everyone else's so-called rules. The OG Ro5 formulation is probabilistic in nature. It defines a space where oral absorption is most likely. It's thus also highly prone to exceptions. 2/
From the beginning, the Ro5 has been challenged. Chemists are quick to point to all the marketed drugs that fall outside of Ro5 space. Behold venetoclax (below). Behold pretty much every oral macrolide drug there is. 3/
In fact, behold roughly half of all marketed oral drugs. (Assuming the rule was applied correctly here, see next post.) 4/
It's also prone to misapplication. As Bob @med_chemist notes in this thread, the Ro5 is only broken if >1 of the metrics is out-of-range. One violation does not violate the whole rule. 5/
Bob the Grumpy Med Chemist @med_chemist
Folks have begun systematically analyzing the space outside the Ro5 space. Jan Kihlberg's group coined the beyond Rule of 5 (bRo5) and also extended Rule of 5 (eRo5) acronyms and explored this space in the seminal paper below, and others that followed. 6/
eRo5 is the "natural tail" of the Ro5 distribution; just more improbable. bRo5 are the true rule-breakers that are off in their own property space. Collectively they describe a space where oral absorption is possible, rather than probable (paraphrasing Kihlberg). 7/
PROTACs and other bifunctionals mostly inhabit this larger realm. At @ArvinasInc, we've developed the Arvinas Rules (unpublished & proprietary, sorry) for oral bioavailability (also BBB penetration) as part of our PROTAC Discovery Engine. 9/
I tell medicinal chemists on their first day at Arvinas (also at the interview!), chuck that Ro5 playbook you learned in the trash. Starting today, you're writing a new chapter in a new book. It's one of the things that makes working in this space as a chemist so much fun. 10/
This all fits in with yesterday's post about the low-probability nature of drug discovery. If you take the attitude that PROTACs will never have oral F because they violate the Ro5, you'd never have embarked on the journey. And you'd be mistaken too. /11
Keith Hornberger @KRHornberger
