Wednesday, September 6, 2017

The Readout by Damian Garde & Meghana Keshavan

Welcome to The Readout, where we keep you on top of the latest in biotech. For more in-depth coverage of biopharma, subscribe to STAT Plus. On Twitter: @damiangarde@megkesh, and @statnews.


The many travails of IBM Watson

IBM has been promoting its Watson supercomputer as the key to unraveling complex biology and revolutionizing cancer care. 

A STAT investigation shows that, three years in, it’s not even close. Far from finding effective therapies for cancer, Watson for Oncology is still trying to parse all the different varieties of the disease. And only a few dozen hospitals around the world are working with Watson — a far cry from IBM’s plans for market dominance. Why?

Interviews suggest that IBM, in an effort to buoy flagging revenue, has rolled out an incomplete product that doesn’t suit hospital needs.

Read more.

Venter's genomic facial recognition paper under fire


A striking resemblance? (yaniv erlich) 
Can genomics wizards make an informed guesstimate on what a person looks like, based on his or her DNA? J. Craig Venter sure thinks so, per a new PNAS paper. Yet his Human Longevity Institute study is facing Twitter blowback for that claim. 

Former employee Jason Piper is listed as an author on the study, but he takes exception with its conclusion — and maintains that the methodology simply generates images that look like generic versions of any person’s given race and gender. 

Yaniv Erlich, a computer scientist at Columbia University — and, notably, chief scientific officer of rival DNA analysis company MyHeritage —  has similar complaints. When Venter generated an image of his own face based on his genes, for instance, Erlich said it looked an awful lot like like actor Bradley Cooper (see above). He reviewed the Venter team's paper for Science and rejected it; he told STAT that he found the methodology severely flawed, even worrisome. Later today, he plans to publish his own brief explaining his concerns.

Why is this one study causing such a fuss? Venter's name, for one; he's a larger-than-life and perpetually controversial figure in the field of genomics. Then there's the topic itself. Predicting a person’s physical appearance from just their DNA could be a powerful tool for criminal investigations. And it could have far-reaching implications for personal privacy.

“This is the thing: It’s dangerous, not fun,” Erlich said. “I was really upset about this paper, and am trying to undo the damage it could cause.” 

Human Longevity is rebutting these criticisms, issuing a statement declaring that the researchers (sans Piper, clearly) "stand behind their methodology."

The facial recognition research was conducted, HLI says, in part to point out that "individuals who are participating in these studies need to fully understand the implications of having their genomes in such databases."

Sponsor content by BD Genomics

BD Genomics

BD is directing its legacy of innovation and quality to advance genomics research, offering high-value products designed to deliver greater efficiency and precise results behind every scientific discovery. Your search for answers in genomics can depend on a difference of one — one genetic variant, one rare transcript, one specialized cell’s profile. It all comes down to the difference between insight and information loss. Discover the difference one company can make. Discover the new BD.

An award for TOR

This year's Lasker Awards are out, and among the honorees is the University of Basel's Michael Hall for his role in demystifying the TOR pathway. 

Tinkering with TOR has implications for a host of diseases, including cancer and diabetes, and biotech companies like Navitor and Alector have raised cash and hopes alike with plans to craft drugs against the target. 

The other biomedical winners are National Cancer Institute researchers Douglas R. Lowy and John T. Schiller, who helped bring about HPV vaccines. Planned Parenthood took the public service honor. 

Back to school? We've got your cheat sheets

In the back-to-school spirit, we've put together a slew of cheat sheets for STAT Plus subscribers: Quick, handy guides to key terms in biopharma. We’re talking acronyms like CRISPR and MACRA, lingo like p value and biosimilars, and technologies like optogenetics and gene drives. Browse our list here and let us know here what you think we should add to the resource library.

Insmed gets on the board with a September success


Jennifer keefe/STAT

It’s week one in biotech’s September of debatable content, and one major binary came up positive: Insmed’s inhaled antibiotic met its goals in a phase 3 trial.

The company’s drug, dubbed ALIS, is meant to treat a rare lung disease caused by bacteria. In the pivotal study, ALIS completely cleared the offending bacteria in 29 percent of patients, handily beating the 9 percent seen on placebo. Now Insmed is plotting an FDA submission, news that nearly doubled its share price Tuesday.

More reads

  • Biotech CEOs stand up for DACA in an open letter. (Read it here)
  • A new piece in the Parkinson's puzzle. (Derek Lowe)
  • Genetic insights are helping scientists navigate toward neurological disease treatments. (C&EN)
  • New blood needed: Pharma R&D leadership tenure. (LifeSciVC)

Have a news tip or comment you want to send us?

Send us an email

Thanks for reading! Until tomorrow,

Damian & Meghana

Enjoy this email? Tell your friends and coworkers to sign up here.