The FDA has turned down Bio­haven’s drug for spin­ocere­bel­lar atax­ia, a neu­rode­gen­er­a­tive dis­ease that leads to loss of co­or­di­na­tion and bal­ance, cit­ing is­sues with us­ing re­al-world ev­i­dence and nat­ur­al his­to­ry stud­ies. As a re­sult, Bio­haven is cut­ting costs, paus­ing pro­grams and may lay off staff.

Known as tro­r­ilu­zole, it’s the lat­est rare dis­ease drug to be re­ject­ed or face road bumps un­der new FDA lead­er­ship. The chal­lenges have come seem­ing­ly af­ter com­pa­nies and the agency had pre­vi­ous­ly agreed on how a study would be con­duct­ed — elic­it­ing frus­tra­tion from drug de­vel­op­ers. Oth­ers, like UniQure and Replimune, have al­so been turned away by the agency for us­ing tri­als with­out a typ­i­cal con­trol arm that’s em­ployed in ran­dom­ized con­trolled tri­als.

Over the last dozen years, Chris Gib­son has turned Re­cur­sion from a scrap­py Salt Lake City start­up in­to a lead­ing face of the AI-fo­cused biotech world. On Wednes­day, he an­nounced he will step down as CEO at the be­gin­ning of 2026.

Na­jat Khan, a for­mer J&J ex­ec­u­tive who joined Re­cur­sion about 18 months ago as chief R&D and com­mer­cial of­fi­cer, will suc­ceed Gib­son.

Time and again, Gib­son has proven to be a shrewd CEO, es­pe­cial­ly in read­ing the mar­ket's tem­per­a­ture and re­spond­ing ac­cord­ing­ly, from com­plet­ing a flashy IPO in the boom times of 2021, to merg­ing with Ex­sci­en­tia in 2024, to slash­ing its pipeline and work­force in 2025.

Re­search progress on us­ing com­put­ers to de­sign pro­teins like an­ti­bod­ies is still in the ear­ly days, ac­cord­ing to No­bel lau­re­ate David Bak­er.

In a pa­per pub­lished Wednes­day in Na­ture, Bak­er de­tails the work be­hind RFan­ti­body, the dif­fu­sion-based, an­ti­body-mak­ing AI mod­el from his lab. That in­cludes gen­er­at­ing full-length an­ti­bod­ies by com­pu­ta­tion­al de­sign. Bak­er, di­rec­tor of the Uni­ver­si­ty of Wash­ing­ton's In­sti­tute for Pro­tein De­sign, had al­ready shared much of the work in preprints go­ing back to last March. In an in­ter­view with End­points News, he laid out where he sees the field go­ing next.

"This is very much a cul­mi­na­tion, but it's not the end of the sto­ry," Bak­er told End­points. "I think we'll see con­tin­ued im­prove­ment in com­pu­ta­tion­al an­ti­body de­sign, to the point where I ex­pect, with­in sev­er­al years, it will be­come the in­dus­try stan­dard."