NYU Data Science newsletter – January 13, 2016

GROUP CURATION: N/A

… Before researchers start clinical trials, they’re supposed to pre-specify which health outcomes they’re most interested in. For an antidepressant, these might include people’s self-reports on their mood, how the drug affects sleep, sexual desire, and even suicidal thoughts.

The idea is that researchers won’t just publish positive or more favorable outcomes that turn up during the study, while ignoring or hiding important results that don’t quite turn out as they were hoping.

But that doesn’t always happen. “In Study 329,” explains Ben Goldacre, a crusading British physician and author, “none of the pre-specified analyses yielded a positive result for GSK’s drug, but a few of the additional outcomes that were measured did, and those were reported in the academic paper on the trial, while the pre-specified outcomes were dropped.”

As an assistant professor at Princeton in 1999, the political scientist Kathleen McNamara co-wrote an article for Foreign Affairs urging more democratic oversight of the European Central Bank. The department reprimanded her for it. Junior faculty, like children in Victorian novels, were to be seen, not heard.

Times have changed. Universities now generally encourage efforts to reach a broad public, although those efforts lend, as one scholar put it, more “warm atmospherics” than significant heft to hiring, promotion, and tenure portfolios. High-profile blogs like The Monkey Cage, started by George Washington University’s John Sides in 2007 and adopted by The Washington Post in 2013, have offered a platform for such outreach, bolstering political scientists’ sense of relevance and public engagement. The blogs, while influenced by political journalism’s hectic daily tempo, have also influenced that journalism, pressuring reporters to incorporate data-driven theory into their reporting to help contextualize the chaotic ups and downs of elections and wars, treaties and massacres.

Datadog, a company with a cloud service for monitoring the performance of applications and their underlying infrastructure, is announcing today a $94.5 million round of funding — a hefty pile of money considering that the company revealed a $31 million round less than a year ago.

Datadog offers developers and operations staffers a way to keep track of how everything in their tech stack is doing. It can provide dashboards, send alerts, and store information on who fixes bugs and how.

It is a movement building steady momentum: a call to make research data, software code and experimental methods publicly available and transparent. A spirit of openness is gaining traction in the science community, and is the only way, say advocates, to address a ‘crisis’ in science whereby too few findings are successfully reproduced. Furthermore, they say, it is the best way for researchers to gather the range of observations that are necessary to speed up discoveries or to identify large-scale trends.

The open-data shift poses a conundrum for junior researchers, who are carving out their niche. On the one hand, the drive to share is gathering official steam. Since 2013, global scientific bodies — including the European Commission, the US Office of Science and Technology Policy and the Global Research Council — have begun to back policies that support increased public access to research.

Countless therapeutics – and certainly our earliest ones – were discovered from natural sources. Lodo Therapeutics, a new small molecule drug discovery startup based in New York City, is basing its platform on this premise: It will develop medicines that can be produced by uncultured soil bacteria and the human microbiome, chasing global health indications like tuberculosis. … “More than half of all small molecule drugs for cancer, infections and Type 2 diabetes today are derived from natural products, representing significant promise of this approach for patients,” Brady said in a statement. “Our genome-based, culture-independent approach exploits the power of microbial evolution to identify therapeutically valuable natural products.”

… Philip Purnell at the University of Leeds, UK, envisions a city where everyday maintenance issues will not only be reported by mobile robotic sensors roaming the metropolis, but one where the repairs themselves will be carried out by robots. In what Purnell and colleagues call “self-repairing cities”, a robot could get to work on a problem as soon as it is reported.

Rob Richardson, a colleague of Purnell’s also at the University of Leeds, shares the enthusiasm. He thinks on-site robots carrying out precision repairs will avoid the need for disruptive construction jobs in the heart of our cities.

… “What we’ve developed is a visual recognition system that will identify the patient’s face, identify the medication they’re taking, confirm ingestion, and then confer that data back to the care provider or to a pharmaceutical company conducting a clinical trial,” AiCure CEO Adam Hanina told MobiHealthNews.

Hanina says the app can run on any mobile device with a camera. It doesn’t actually record video or take photos, it just uses the phone’s camera coupled with image recognition technology to confirm a patient has taken their medication and reports that to the patient’s care provider or the organizer of a clinical trial. Depending on the use case, AiCure can also build in motivational tools or other ways to promote adherence.

The German car manufacturer admitted cheating emissions tests. Which companies are impacted directly or indirectly by these revelations? James Phare from Data To Value used graph analysis and open source information to unravel the impact of the VW scandal on its customers, partners and shareholders.

… Given that many brains, mine included, have a hard time grasping Bayes’ theorem, the Bayesian-brain thesis might seem surprising–and in fact it has provoked pushback. Seeking insight into the debate, last month my brain and I attended a two-day meeting at New York University, “Is the Brain Bayesian?”

The meeting was organized by philosophers Ned Block and David Chalmers of the NYU Center for Mind, Brain and Consciousness. The Center has been busy. In November, it sponsored a workshop on integrated information theory, which I critique here. Whereas integrated information theorists seek to explain consciousness, how the mind feels, Bayesians focus on cognition, what the mind does.

The Field Guide to Data Science spells out what data science is, why it matters to organizations, as well as how to create data science teams. Along the way, our team of experts provides field-tested approaches, personal tips and tricks, and real-life case studies. Senior leaders will walk away with a deeper understanding of the concepts at the heart of data science, practitioners will add to their toolboxes, and beginners will find insights to help them start on their data science journey.

We just finished the second day of a workshop on Docker at the Berkeley Institute for Data Science. I was invited to organize the workshop after some Berkeley folk couldn’t make our Davis workshop in November, and so I trundled on down for two days to give it a try. About 15 people showed up, from all walks of research and engineering. (The materials are freely available under a CC0 license.) I did the intro stuff, and Luiz Irber and Carl Boettiger both taught sections on their more advanced uses of Docker.

tl;dr? I think the workshop went well, in the sense that the attendees all got some hands-on experience with Docker and engaged its potential benefits (as well as some of its limitations).