Ear­li­er this month, Flag­ship an­nounced their big bet on the soft­ware half the in­dus­try is talk­ing about, launch­ing the AI and ma­chine learn­ing start­up. Now, they and a cou­ple oth­er in­vestors are gam­bling $100 mil­lion on a soft­ware that much of the pub­lic gen­er­al­ly thinks of as a cool, IT af­ter­thought: cloud com­put­ing.

The idea, says founder and Flag­ship part­ner David Berry, is one of scale: The sheer mag­ni­tude of bi­o­log­i­cal da­ta that you can store on cloud tech­nol­o­gy is un­prece­dent­ed. And that size, when lever­aged prop­er­ly, can al­low you to ask ques­tions and form in­sights that are sim­i­lar­ly un­prece­dent­ed.

David Berry

“When you start tak­ing off the con­straints, then you can start talk­ing com­pu­ta­tion­al prob­lems that are more in­ter­est­ing,” Berry tells End­points News. “We all talk a lot about the dif­fer­ent method­olo­gies you can use to com­pute, but again, the da­ta is what you’re com­put­ing.”

Hence: Va­lo Health. Like most Flag­ship launch­es, the start­up has been in­cu­bat­ing for some time and al­ready boasts about 100 em­ploy­ees, half work­ing on the mon­i­tors and half work­ing in the lab. With the as­sis­tance of tech­nol­o­gy brought in from Nu­mer­ate and For­ma Ther­a­peu­tics, those re­searchers have cre­at­ed what Va­lo calls the “Opal Com­pu­ta­tion­al Plat­form,” a set of cloud-based tech­nol­o­gy meant to re­think drug de­vel­op­ment from tar­get dis­cov­ery through prod­uct se­lec­tion and all the way down to clin­i­cal tri­al de­sign.

And, like many Flag­ship star­tups, those re­searchers have had a lot of cash to make it hap­pen. With a few un­named in­vestors con­tribut­ing half the pot, they raised a rough­ly $100 mil­lion Se­ries A in 2019 to get start­ed.

It’s still ear­ly days for Va­lo and they’re not dis­clos­ing a few bits of key in­fo, in­clud­ing where they’re get­ting the da­ta they’re loft­ing up to their cloud or what and how many prod­ucts they’ve set­tled on to start.

Still, the am­bi­tion is clear. By an­a­lyz­ing huge amounts of hu­man da­ta, they’ll try to short­cir­cuit the prob­lems with an­i­mal and oth­er mod­els and cut down on the high costs and the fail­ure of drug de­vel­op­ment. Berry knows that this is what most da­ta sci­ence com­pa­nies have been promis­ing for years, with not a ton yet to show for it, but he ar­gues these com­pa­nies have in fact been think­ing too small. Cal­cu­lat­ing the fail­ure rates at var­i­ous stages of de­vel­op­ment from mol­e­cule se­lec­tion to com­mer­cial mar­ket­ing, he came up with a back-of-the-nap­kin suc­cess rate for drugs of about 1 in 4000.

“We see all too of­ten this frame­work where peo­ple are say­ing: What if we could dou­ble the prob­a­bil­i­ty of suc­cess?” Berry says. “And, re­al­ly when you’re deal­ing with that foun­da­tion, isn’t the right ques­tion: How can we rewrite the process.”

Va­lo will look to do that in three ar­eas: car­dio­vas­cu­lar dis­ease, neu­rode­gen­er­a­tion and on­col­o­gy. Al­though car­dio­vas­cu­lar dis­ease re­mains one of the most com­mon caus­es of death in the US, few biotechs go in­to the field these days, large­ly be­cause drugs in this field tend to re­quire huge tri­als and it’s dif­fi­cult to show ben­e­fit above ex­ist­ing drugs.

Berry says, though, that the field is “da­ta-rich,” a prime area for Va­lo’s tech­nol­o­gy to find new ap­proach. And that, he adds, is es­pe­cial­ly true for neu­rode­gen­er­a­tion, a form of dis­ease that is ex­pect­ed to on­ly grow in com­ing years and where tri­al af­ter tri­al has stum­bled.

“Said sim­ply: Hu­mans haven’t been to solve this dis­ease. We think these are com­plex, in­te­grat­ed dis­eases, they need some­thing deep and in­sight­ful to be able to un­lock it,” he says. “We see that as a fron­tier.”