Minkyung Baek (University of Washington)

'They have shown that this is not some im­pos­si­ble thing': Aca­d­e­m­ic lab copies Google’s big bi­o­log­i­cal break­through

When Demis Has­s­abis, CEO of Google’s AI out­fit Deep­Mind, an­nounced last year that they had cracked one of the tough­est puz­zles in bi­ol­o­gy — suc­cess­ful­ly pre­dict­ing a pro­tein’s shape from its amino acid se­quence — Minkyung Baek watched with a cu­ri­ous mix­ture of dread and ex­cite­ment.

“It felt like I just lost my job,” said Baek, a post­doc at the Uni­ver­si­ty of Wash­ing­ton’s In­sti­tute for Pro­tein De­sign.

Baek had spent the last three years try­ing to do the ex­act same thing, work­ing with what for years had been the lead­ing pro­tein de­sign tech­nol­o­gy at the lead­ing pro­tein de­sign lab. Now Deep­Mind had ful­ly eclipsed her. She felt the sci­en­tif­ic com­mu­ni­ty’s ex­cite­ment, even as she pon­dered her pro­fes­sion­al fu­ture.

Deep­Mind’s break­through, though, had a caveat: The com­pa­ny hadn’t ac­tu­al­ly shown how they cracked the puz­zle, nor had they made their head­line-grab­bing soft­ware avail­able to any re­searchers out­side of Google’s dis­parate of­fices. Baek won­dered if she could, with the few bread­crumbs Google left, re­con­struct Deep­Mind’s soft­ware and dis­trib­ute it to the world.

On Thurs­day, she showed she could. Baek and her team at UW’s Bak­er lab de­tailed ma­chine learn­ing soft­ware in Sci­ence that was al­most — al­beit not quite — as pow­er­ful as Deep­Mind in pre­dict­ing a pro­tein’s struc­ture from its se­quence, and demon­strat­ed how it could be de­ployed to probe ques­tions in­tri­cate­ly linked to un­der­stand­ing dis­ease and de­sign­ing new drugs. They have made the tool, known as RoseTTaFold, avail­able on GitHub, where UW claims it has al­ready been down­loaded by over 140 re­search teams.

John Moult

“If you want­ed to be neg­a­tive about it, you could say they’re play­ing catch-up and got re­sults that were not quite as good,” said John Moult, a com­pu­ta­tion­al bi­ol­o­gist at the Uni­ver­si­ty of Mary­land who in 1994 launched the an­nu­al chal­lenge where Deep­Mind de­buted their re­sults. “I think the more pos­i­tive and prop­er way of look­ing at it is that they have done it near­ly as well and they have al­ready pro­vid­ed a serv­er, which works at least in a cou­ple of times. And they have done a full re­lease of their code.

“They have shown that this is not some im­pos­si­ble thing for oth­er peo­ple to achieve,” Moult added.

The pa­per came out the same time Na­ture pub­lished the de­tailed meth­ods be­hind Deep­Mind’s work — a co­in­ci­dence the folks at the Bak­er lab chalk up to their de­ci­sion to re­lease the preprint and open source soft­ware last month. Has­s­abis’ team sim­i­lar­ly promised to make their soft­ware pub­lic, al­though they have not pro­vid­ed a sim­i­lar serv­er to Baek’s.

The two ap­proach­es are broad­ly sim­i­lar on a high lev­el, said Baek, re­ly­ing on broad datasets of known pro­tein se­quences and struc­tures and sim­i­lar­i­ties be­tween co-evolved pro­teins. But they dif­fer vast­ly in how they tech­ni­cal­ly car­ry out their vi­sion.

Col­lec­tive­ly, they give re­searchers two so­lu­tions to a decades-old prob­lem and of­fer im­prove­ments over the orig­i­nal re­sults re­leased last year, in­clud­ing the abil­i­ty to pre­dict struc­ture in min­utes or hours, rather than days.

A pro­tein’s func­tion is de­ter­mined by its struc­ture, but for decades the on­ly way to de­ter­mine that struc­ture was through a va­ri­ety of imag­ing tech­niques, such as X-ray crys­tal­log­ra­phy, that could be lengthy or dif­fi­cult and didn’t pro­duce ac­cu­rate de­pic­tions for every pro­tein. Al­though those tech­niques have im­proved over time, the po­ten­tial to pre­dict a a struc­ture from its se­quence alone — known as the fold­ing prob­lem — con­tin­ued to be a holy grail.

David Bak­er

Baek’s lab, run by bio­chemist David Bak­er and fa­mous for its work in de­sign­ing pro­teins from scratch, had been lead­ing the race for years be­fore Deep­Mind leapfrogged them. See­ing the com­pa­ny’s progress, they adopt­ed more ma­chine learn­ing tech­niques to catch up.

In their pa­per, Baek showed a few of the ways wide­spread ac­cess to such tech­nol­o­gy could be used to build new drugs. They pre­dict­ed struc­tures for three class­es of pro­teins key to a range of dis­eases, in­clud­ing can­cer and de­men­tia, show­ing how rare mu­ta­tions could warp the pro­tein’s shape and re­veal­ing pos­si­ble open­ings to tar­get drugs.

“It’s not as good as Deep­Mind,” said Jin­bo Xu, a com­pu­ta­tion­al bi­ol­o­gist at the Uni­ver­si­ty of Chica­go. “But some of the struc­tures are very ac­cu­rate. It will be use­ful.”

Un­til Deep­Mind re­leas­es their open-source work, RoseTTaFold, Xu said, will be more help­ful to the field. They may not have long to wait, though.

Baek post­ed their pa­per as a preprint on June 15. Three days lat­er, Has­s­abis tweet­ed out a “brief up­date” on Deep­Mind’s pro­tein fold­ing soft­ware, say­ing that a full pa­per out­lin­ing their meth­ods was un­der re­view and that they would be pro­vid­ing the source code and “broad ac­cess” to the soft­ware to re­searchers. Xu said a serv­er for re­searchers to ac­cess Deep­Mind’s tech would be the ide­al tool.

Moult not­ed the Bak­er pa­per took their soft­ware to places that Deep­Mind didn’t. In ad­di­tion to pre­dict­ing the struc­ture of in­di­vid­ual pro­teins, they al­so pre­dict­ed how dif­fer­ent pro­teins come to­geth­er to form larg­er struc­tures, as they do to per­form man­i­fold bi­o­log­i­cal func­tions. Baek pre­dict­ed the dif­fer­ences be­tween two im­mune com­plex­es: one formed by IL-23 and one by IL-12. The re­sults could help drug de­vel­op­ers iden­ti­fy mol­e­cules that block one but not the oth­er, cre­at­ing more pre­cise drugs for au­toim­mune dis­eases.

Baek said her next big project is im­prov­ing the soft­ware’s abil­i­ty to pre­dict these in­ter­ac­tions.

Biotech Half­time Re­port: Af­ter a bumpy year, is biotech ready to re­bound?

The biotech sector has come down firmly from the highs of February as negative sentiment takes hold. The sector had a major boost of optimism from the success of the COVID-19 vaccines, making investors keenly aware of the potential of biopharma R&D engines. But from early this year, clinical trial, regulatory and access setbacks have reminded investors of the sector’s inherent risks.

RBC Capital Markets recently surveyed investors to take the temperature of the market, a mix of specialists/generalists and long-only/ long-short investment strategies. Heading into the second half of the year, investors mostly see the sector as undervalued (49%), a large change from the first half of the year when only 20% rated it as undervalued. Around 41% of investors now believe that biotech will underperform the S&P500 in the second half of 2021. Despite that view, 54% plan to maintain their position in the market and 41% still plan to increase their holdings.

Sur­geons suc­cess­ful­ly at­tach pig kid­ney to a hu­man for the first time, us­ing tech from Unit­ed's Re­vivi­cor

In a first, researchers reportedly successfully transplanted a pig kidney into a human without triggering an immediate immune response this week. And the technology came from the biotech United Therapeutics.

Surgeons spent three days attaching the kidney to the patient’s blood vessels, but when all was said and done, the kidney appeared to be functioning normally in early testing, Reuters and the New York Times were among those to report. The kidney came from a genetically altered pig developed through United’s Revivicor unit.

Endpoints News

Keep reading Endpoints with a free subscription

Unlock this story instantly and join 120,100+ biopharma pros reading Endpoints daily — and it's free.

David Livingston (Credit: Michael Sazel for CeMM)

Renowned Dana-Far­ber sci­en­tist, men­tor and bio­phar­ma ad­vi­sor David Liv­ingston has died

David Livingston, the Dana-Farber/Harvard Med scientist who helped shine a light on some of the key molecular drivers of breast and ovarian cancer, died unexpectedly last Sunday.

One of the senior leaders at Dana-Farber during his nearly half century of work there, Livingston was credited with shedding light on the genes that regulate cell growth, with insights into inherited BRCA1 and BRCA2 mutations that helped lay the scientific foundation for targeted therapies and earlier detection that have transformed the field.

No­vo CEO Lars Fruer­gaard Jør­gensen on R&D risk, the deal strat­e­gy and tar­gets for gen­der di­ver­si­ty


I kicked off our European R&D summit last week with a conversation involving Novo Nordisk CEO Lars Fruergaard Jørgensen. Novo is aiming to launch a new era of obesity management with a new approval for semaglutide. And Jørgensen had a lot to say about what comes next in R&D, how they manage risk and gender diversity targets at the trendsetting European pharma giant.

John Carroll: I’m here with Lars Jørgensen, the CEO of Novo Nordisk. Lars, it’s been a really interesting year so far with Novo Nordisk, right? You’ve projected a new era of growing sales. You’ve been able to expand on the GLP-1 franchise that was already well established in diabetes now going into obesity. And I think a tremendous number of people are really interested in how that’s working out. You have forecast a growing amount of sales. We don’t know specifically how that might play out. I know a lot of the analysts have different ideas, how those numbers might play out, but that we are in fact embarking on a new era for Novo Nordisk in terms of what the company’s capable of doing and what it’s able to do and what it wants to do. And I wanted to start off by asking you about obesity in particular. Semaglutide has been approved in the United States for obesity. It’s an area of R&D that’s been very troubled for decades. There have been weight loss drugs that have come along. They’ve attracted a lot of attention, but they haven’t actually ever gained traction in the market. My first question is what’s different this time about obesity? What is different about this drug and why do you expect it to work now whereas previous drugs haven’t?

Endpoints Premium

Premium subscription required

Unlock this article along with other benefits by subscribing to one of our paid plans.

Man­u­fac­tur­ing woes for No­vavax’s Covid jab bad­ly dis­rupt plans for roll­out to the poor — re­port

Production problems at a Novavax facility in Maryland have led to delays in the Covax vaccine sharing program. Now, a shortage of 1 billion doses is expected, as the supplier tries to navigate producing a shot up to regulators’ standards, Politico reported Tuesday.

The company has run into trouble with the purity of the vaccine. Novavax has had trouble proving it can produce a shot consistently up to standards, and it has caused significant delays in the rollout to low- and middle-income countries. This follows several delays at Novavax that has put the executive crew on the defensive.

Break­ing: Bio­gen sells just $300K worth of Aduhelm in Q3, as ques­tions on long term vi­a­bil­i­ty re­main

Barely anyone is accessing Biogen’s controversial Alzheimer’s treatment, with the company reporting just $0.3 million in Aduhelm sales in the third quarter. Although investors will be looking to the longer term, when CMS may decide to cover the drug and open the floodgates for more coverage, use of the drug is currently stalled.

Since June, when the FDA first signed off on the drug under its accelerated pathway, Biogen said Wednesday that it’s sold a total of $2 million worth of Aduhelm. Biogen said on its earnings call that about 120 sites so far have infused at least one patient with Aduhelm, which is priced at $56,000 annually. Morgan Stanley previously predicted about 14,000 patients will access Aduhelm by the end of 2022.

Endpoints News

Keep reading Endpoints with a free subscription

Unlock this story instantly and join 120,100+ biopharma pros reading Endpoints daily — and it's free.

Bill Gates at the Global Investment Summit in London, Oct. 19, 2021 (Leon Neal/Pool via AP Images)

Gates Foun­da­tion pledges $120M to ramp up gener­ic sup­ply of Mer­ck­'s Covid-19 pill while ac­tivists blast Pfiz­er's dis­pro­por­tion­ate pow­er

Merck’s molnupiravir may not be officially authorized anywhere in the world yet, but who will get access to it has shaped up to be a huge issue. The Bill & Melinda Gates Foundation is now stepping up to ensure lower-income countries won’t be left behind — and calling on others to follow its lead.

The oral antiviral pill, which was shown to dramatically cut the risk of severe Covid-19 disease and death in a Phase III study, is the latest rallying symbol in the battle against not just the coronavirus but the inequality it’s exposed.

With hun­dreds of mil­lions spent on failed ac­cel­er­at­ed ap­provals, re­searchers call for faster FDA with­drawals

Between 2017 and 2019, Medicare spent more than $220 million on cancer drugs for which the indications were either voluntarily pulled by their applicants or FDA’s oncology adcomm had recommended their withdrawal.

That kind of massive spending on cancer drugs lacking overall survival benefit is wasteful and risks harming people’s health, a research letter published in JAMA Internal Medicine on Monday said. The researchers from Harvard and the London School of Economics called on the FDA to move faster in both requiring timely postmarketing trials and accelerating the speed in pulling these dangling approvals when the confirmatory studies fail.

Endpoints News

Keep reading Endpoints with a free subscription

Unlock this story instantly and join 120,100+ biopharma pros reading Endpoints daily — and it's free.

Raymond Stevens, ShouTi Pharma CEO

A new Schrödinger-backed start­up emerges from the sci­en­tist who mapped the first hu­man GPCR

One of the most popular targets in drug development, representing about a third of existing drugs, are G-protein coupled receptors — the tiny but integral membrane proteins responsible for recognizing things like light, taste, smell, hormones and pain.

But due to challenges in mapping their structure, the protein family remains largely unexplored.

A slate of companies has emerged over the last few years to change that. If one can figure out the structure of these elusive membrane receptors, it might be possible to create small molecule drugs that overcome the limitations of, say, biologic and peptide therapies. That promise is what gets serial entrepreneur Raymond Stevens out of bed in the morning.