Less than a week af­ter Roche joined forces with Dyno Ther­a­peu­tics to de­vel­op gene ther­a­pies us­ing ar­ti­fi­cial in­tel­li­gence, its gi­ant sub­sidiary Genen­tech is hop­ping on the AI band­wag­on with a dif­fer­ent play­er.

Genen­tech has inked a deal with Stan­ford spin­out Gen­e­sis Ther­a­peu­tics to har­ness its AI pow­er for drug de­vel­op­ment and dis­cov­ery. Gen­e­sis is get­ting an up­front pay­ment and mile­stones, but the com­pa­nies are keep­ing the de­tails un­der wraps for now. The Burlingame, CA-based biotech al­so stands to earn fu­ture roy­al­ties on any ap­proved Genen­tech drugs that come from the deal.

Us­ing AI tech­nol­o­gy, Gen­e­sis is able to make “ul­tra-fast and ac­cu­rate pre­dic­tions” of a com­pound’s po­ten­cy, se­lec­tiv­i­ty, tox­i­c­i­ty and more, CEO Evan Fein­berg told End­points News. “That helps us get them pre­pared for clin­i­cal tri­als more quick­ly, and with a more op­ti­mal com­pound,” he said.

Evan Fein­berg

“We are screen­ing any­where from mil­lions to bil­lions of com­pounds in sil­i­co at each stage, from hit iden­ti­fi­ca­tion through hit-to-lead, lead op­ti­miza­tion and can­di­date se­lec­tion,” he added lat­er.

The biotech was formed out of Stan­ford Uni­ver­si­ty’s Pande Lab in 2019, where Fein­berg co-in­vent­ed Po­ten­tial­Net — a neur­al net­work de­signed to pre­dict pro­tein−lig­and bind­ing and mol­e­c­u­lar prop­er­ties. The com­pa­ny is backed by An­dreessen Horowitz, and touts Ami­ra Phar­ma­ceu­ti­cals’ founder and long­time Ver­sant ad­vi­sor Pep­pi Pr­a­sit as its act­ing CSO, with Alex­ion founder Leonard Bell along for the ride as chair­man of the board.

Gen­e­sis seeks to use AI to de­vel­op can­di­dates with “su­pe­ri­or se­lec­tiv­i­ty,” thus lim­it­ing side ef­fects, which Fein­berg said “is what this is all about.” The CEO said he’s had four sig­nif­i­cant leg surg­eries in the last 10 years, one of which left a pe­riph­er­al nerve per­ma­nent­ly dam­aged.

“I’ve had first-hand ex­pe­ri­ence with most or many neu­ro­mus­cu­lar, mus­cu­loskele­tal-re­lat­ed drugs that have re­al­ly bad side ef­fects. So I… feel very per­son­al­ly com­mit­ted to cre­at­ing a phar­ma­copoeia that has bet­ter qual­i­ty of life for pa­tients,” he said.

Pep­pi Pr­a­sit

The deal ar­rives right on the heels of Aviv Regev’s ar­rival as the new head of re­search at gRED. Regev has been a star at the Broad, work­ing in her high­ly spe­cial­ized field of com­pu­ta­tion­al bi­ol­o­gy. And she’s ex­pect­ed to play a big role at Genen­tech adding to their strengths in ar­ti­fi­cial in­tel­li­gence and ma­chine learn­ing.

In its col­lab­o­ra­tion with Dyno, Roche’s Spark team will work on im­prov­ing AAV vec­tors. Dyno de­signs, tests and val­i­dates the vec­tors, while part­ners add their pay­loads and work on new de­vel­op­ment pro­grams. James Sabry, glob­al head of Roche Phar­ma Part­ner­ing, said the deal will “bring us added abil­i­ty to go to AAV 2.0.” AI and ma­chine learn­ing, he added, are “no longer things of the fu­ture when it comes to drug dis­cov­ery.”

“AI can help un­lock the next gen­er­a­tion of in­no­v­a­tive ther­a­pies for pa­tients in need of ad­di­tion­al op­tions. We are ex­cit­ed to work with Gen­e­sis’ team to dis­cov­er med­i­cines cur­rent­ly out of reach us­ing con­ven­tion­al meth­ods,” Sabry said in a state­ment.

Dupixent, the blockbuster anti-inflammatory drug from Sanofi and Regeneron, has cleared a high-stakes Phase III study in chronic obstructive pulmonary disease, the companies announced Thursday morning.

If they hold up in a second, identical trial, the data pave the way for Dupixent to become the first biologic to treat patients whose COPD remains uncontrolled despite being on maximal standard-of-care inhaled therapy — the patient population studied in the pivotal program. The companies had spotlighted this as a key readout as they look to expand the Dupixent franchise and explore its full potential.

Incyte touted an accelerated approval for its PD-1 retifanlimab in a rare skin cancer on Wednesday, roughly a year and a half after the drug suffered a rejection in squamous cell carcinoma of the anal canal (SCAC).

Retifanlimab, marketed as Zynyz, was approved for metastatic or recurrent locally advanced Merkel cell carcinoma (MCC), a fast-growing skin cancer typically characterized by a single, painless nodule. It’s roughly 40 times rarer than melanoma, according to the nonprofit Skin Cancer Foundation — but incidence is growing, particularly among older adults, Incyte said in its announcement.

Novartis is calling off a UK-based trial for Leqvio in the primary prevention of cardiovascular events in patients with high cholesterol, the company confirmed on Wednesday.

The Swiss pharma giant made the decision after “careful evaluation,” a spokesperson told Endpoints News via email. The trial, dubbed ORION-17, was planned in partnership with England’s National Health Service (NHS) and was part of the company’s strategy to establish Leqvio as a standard of care in cardiovascular disease management.