As more and more bio­phar­mas de­vel­op ar­ti­fi­cial in­tel­li­gence plat­forms, the drug dis­cov­ery process is be­ing re­shaped to fit new goals on cut­ting down the prodi­gious amount of time, en­er­gy and mon­ey that go in­to a drug pro­gram. Now one of the most am­bi­tious play­ers in the dri­ve to im­prove on ROI, As­traZeneca, is mark­ing a mile­stone on that front by adding the first tar­get gen­er­at­ed by AI to its port­fo­lio.

The tar­get comes out of a col­lab­o­ra­tion with Benev­o­len­tAI, a Lon­don-based com­pa­ny that first start­ed work­ing with the big phar­ma back in April, 2019. Thanks to the biotech’s plat­form AI tech, As­traZeneca has iden­ti­fied and val­i­dat­ed a new way to at­tack chron­ic kid­ney dis­ease and will be­gin de­vel­op­ing com­pounds cen­tered on this tar­get.

This se­lec­tion fits in­to a broad­er AI strat­e­gy at As­traZeneca, the com­pa­ny’s re­nal bio­sciences chief told End­points News. And the plan isn’t to just lim­it its ap­pli­ca­tions to R&D.

“We’re in­vest­ing in AI and see it as a clear tool that can sup­port our de­ci­sions in drug dis­cov­ery,” Pernille Hansen said. “They can be not on­ly for dis­cov­er­ing new tar­gets, it can be in chem­istry, imag­ing and so forth. So there’s many, many pos­si­bil­i­ties to work with AI.”

But they’re still a long way out from prov­ing AI can of­fer tan­gi­ble ev­i­dence of an im­pact.

Pernille Hansen

Hansen de­clined to say what specif­i­cal­ly the tar­get is, what kinds of drugs will uti­lize it to treat CKD or the time­line for such drug de­vel­op­ment. The on­ly thing she’d dis­close is that it suits their re­nal and AI ap­proach­es, and that As­traZeneca re­searchers have giv­en the OK to ad­vance.

Kid­ney dis­ease is a fa­mil­iar area for the com­pa­ny, giv­en its stud­ies to ex­pand its block­buster Farx­i­ga drug in­to CKD over the last sev­er­al years. A Phase III tri­al in the dis­ease was halt­ed in March over what As­traZeneca deemed “over­whelm­ing ef­fi­ca­cy,” and the com­pa­ny re­leased topline re­sults in Ju­ly. The FDA has set a rough re­view date for some­time dur­ing the sec­ond quar­ter.

If Farx­i­ga does cross the fin­ish line in CKD, it would rep­re­sent a sec­ond ap­proval once thought un­like­ly for a drug class de­vel­oped for di­a­betes. The drug was ap­proved in May to re­duce the risk of car­dio­vas­cu­lar death or hos­pi­tal­iza­tion in heart fail­ure pa­tients with a re­duced ejec­tion frac­tion, and those with or with­out type 2 di­a­betes.

Ivan Grif­fin

Wednes­day’s tar­get, though, comes out of Benev­o­len­tAI’s plat­form, which es­sen­tial­ly forms the shape of a “knowl­edge graph,” COO Ivan Grif­fin told End­points News. Grif­fin’s com­pa­ny has spent years feed­ing its tech with all sorts of da­ta, from pro­teins and genes to re­sults pub­lished in sci­en­tif­ic jour­nals, and train­ing its al­go­rithm to make con­nec­tions that sci­en­tists may not have no­ticed at first.

One could vi­su­al­ize such the graph as an in­ter­con­nect­ed web of “nodes and edges,” Grif­fin said. Sci­en­tists take ad­van­tage of the AI-pre­dict­ed re­la­tion­ships and then in­ter­ro­gate them to see if it holds up, which is what hap­pened in As­traZeneca’s case.

“The strengths we have in the da­ta and AI and ma­chine learn­ing tech­nol­o­gy ap­proach match­es re­al­ly nice­ly, in this case, with what As­traZeneca was able to bring,” Grif­fin said. “By com­bin­ing the two, and then set­ting off to­geth­er to try to dis­cov­er tar­gets and drugs, that was the phi­los­o­phy be­hind start­ing the col­lab­o­ra­tion, and it’s very much felt like that all the way through.

In ad­di­tion to CKD, Benev­o­len­tAI and As­traZeneca are work­ing on de­vel­op­ing drugs for id­io­path­ic pul­monary fi­bro­sis as part of their col­lab­o­ra­tion.

DC District Court Judge Christopher Cooper today granted Vanda Pharma’s request to require the FDA to disclose more info on the complete response letter for its sleep disorder drug Hetlioz.

The melatonin receptor agonist is approved by the FDA to treat non-24-hour sleep-wake disorder, a circadian rhythm disorder. But in 2018 Vanda filed a supplemental application to market Hetlioz as a treatment for jet lag, which the FDA rejected in August 2019, with few details on what Vanda needed to correct course, according to the company.

Biotech Intra-Cellular Therapies touting new data Tuesday for its drug designed to treat major depressive disorder.

The biotech shared results from its Study 403 investigating lumateperone, or Caplyta, as a monotherapy to treat major depressive episodes. The study looked at major depressive disorder (MDD) and bipolar depression with “mixed features.” The trial had three populations: participants with both MDD and bipolar depression, patients with only MDD and patients with only bipolar depression.

Viking Therapeutics will enter the next phase of the GLP-1 weight loss race, as the small biotech touted early data showing its investigational drug led to a drop of as much as 6% more body weight than a placebo.

With research still in the beginning stages of human testing, Viking’s drug will require more fleshed-out data in a Phase II clinical trial, which the San Diego biotech said will start in the middle of this year. Viking hopes the data will help carve a slice of a market dominated by Novo Nordisk and Eli Lilly, with Amgen on their heels and Altimmune trying to push forward despite displeasing investors last week. The companies are developing drugs that have caused a frenzy on social media, in Hollywood and in broader conversations about weight loss drugs and their place in society.