SciVIBE

Pods of Science | Episode 7 | How to Stay Safe from Biothreats JW: Welcome. I’m your host, Jess Wisse. Today I want to share something a little different with you.Let me introduce you to my friend Nick Hennen. He’ll be co-hosting of Pods of Science for us today. This episode was recorded live at the 2020 AAAS meeting. Take it away, Nick.NH: I’m Nick Hennen, Media Relations Advisor for the Pacific Northwest National Laboratory. And I’m here today with Katrina Waters who represents the Biological Division of our laboratory and Kristin Omberg, representing Chemical and Biosignatures Science at PNNL. Today we’ll talk about how increasing globalization is fueling the spread of novel natural biological threats, and advances in biotechnology that could be used to engineer new threats are constantly emerging. Frameworks for assessing unknown biological agents can enable rapid risk profiling and mitigation. This includes applying novel data analysis methods to host-pathogen interaction data to help predict, at early exposure times, whether a patient can be expected to recover from a disease such as Ebola without major interventions.Please introduce yourself and describe what you do at PNNL and why you do it.KW: Hi. So I’m Katrina Waters. I’m a Biochemist and Laboratory Fellow at PNNL. I manage the basic science organization for biology at the lab and work as a researcher in the area of infectious disease and public health. So the reason that I do it is that I get to work with really awesome people who contribute in a lot of different ways and it’s just been a lot of fun. KO: I’m Kristin Omberg. I am the manager of the Chemical and Biological Signatures Group at PNNL which is in the National Security Directorate. I’m a chemist by training and in 1999 I was doing a post-doc at Los Alamos National Laboratory, which is a sister laboratory. And my post-doc didn’t go very well so I started looking for jobs and I got a couple of offers. One was in accelerator production of Tritium and one was in biothreats—so, looking at preparing a system to detect a biological threat in the future. And I talked to my father who happens to be a Nuclear Engineer at PNNL and he said, “I feel kind of good about that counter terrorism stuff.” So I took the job and I started the job in December of 2000. NH: Oh, wow. That’s wonderful. KO: And since then, the 2001 anthrax attacks on the United States, it’s just been a constant sprint. NH: Thanks, Dad.KO: Yeah, thanks, Dad. NH: Tell us briefly about the nature of biothreats and what does that word mean?KO: That is a really interesting question because the way we use biothreat has actually changed in all of our lifetimes. A lot of people don’t realize that up until 1969, the United States had a bioweapons program. So they weaponized Bacillus anthracis, Yersinia pestis and other human pathogens for use in war. The former Soviet Union also had a biological weapons program and they weaponized many of the same pathogens. So at that time biothreat was used to describe a deliberate act of war by a state program using a weaponized pathogen. In 1975, the Biological Weapons Convention came into force and we became less concerned about a state program, both the Soviet Union and the United States ratified that convention. In the 1990s though, we started becoming concerned about terrorist groups. We started worrying about terrorist groups overseas who started using biological agents, and then we saw that in 2001. So we started being concerned about the biothreat by a state actor, terrorist, or a lone actor. But since 2001, we’ve so many outbreaks of diseases that are zoonotic diseases that jump from animals to humans. We’ve had avian influenza, we’ve seen a couple of rounds of Ebola, SARS, MERS and the current coronavirus. And they’ve all demonstrated that they can really be equally devastating in a globalized world. So, in the last decade we’ve started to worry about emerging disease, as well as the health of other

Welcome. I’m your host, Jess Wisse. On today’s episode we’ll talking about how scientists are taking a new approach to better understand and fight cancer. Stay tuned to learn more.   JW: New research published in Cellshows a never-before look the steps that happen when a woman develops endometrial cancer. This type of cancer affects the uterine lining and it can be deadly. PNNL researchers are using their expertise in mass spectrometry and cancer biology to better target this disease. Meet one of them: Karin RodlandKR: I’m Karin Rodland. I am a PNNL laboratory fellow and I'm one of the lead cancer biologists at PNNL. The main thing that I do is provide expertise about cancer biology to the mass spectrometry group that does proteomics and metabolomics measurements of lots of different tumors.As I got into my 30s and I started really doing this, it's like got my PhD and I thought “Where am I going to postdoc? And what am I going to do in my own research lab?” The number of people I knew who were friends who had cancer—it was just mind-boggling. And I would go to Lions Clubs and Rotary Clubs to do this kind of lay outreach and I would start by saying, “If you or someone you know has had cancer raise your hand.” And every single arm in the room would go up. That's why I do this.JW: Karin has studied cancer biology since the 80s and she’s one of the top experts in the field. KR: KR: I ended up at Oregon Health & Science University as an assistant professor in the mid 80s and I was there for 17 years. And I earned the right to go on sabbatical and I came to Pacific Northwest National Lab to learn proteomics because they were the world's best at proteomics and I thought I was going to need that technology for my research. And I came on sabbatical for a year and I really enjoyed the team research philosophy and culture that they do at PNNL. I was totally impressed with the mass spectrometry technologies and the computational biology expertise and I saw a great opportunity to apply all these capabilities to biomedical research and particularly to cancer.JW: For years doctors and scientists have known that cancer is a genetic disease. Our genes control much of what happens in our bodies, including the way our cells function, grow, and divide. Because of this, cancer researchers have spent a lot of time studying the DNA and RNA of cancer cells. But Karin and her team looked one step closer. They studied the proteins synthesized by these cells. KR: What we call the central dogma of molecular biology, is you have the genes and they're the blueprint. And they send out kind of a Xerox copy and that's the messenger RNA. And then the message gets made into proteins and the proteins actually do the work. And the way that they do the work is by being modified with phosphorylation which turns them on and off, or by acetylation which opens them up or closes them down. So it's easy to measure DNA and it's easy to measure RNA. The technology has been very well developed, it's inexpensive, and it's easy. So scientists and doctors measure what is easy and convenient to measure. So we can measure genes and that's the only tool that most doctors have for doing precision medicine. It's the genes and maybe they can do the RNA. And so all these models have been built up trying to predict disease outcome based on the genes and the RNA. What we found is that when you add the proteins you get much, much more information. And sometimes the information from the RNA is a little bit misleading. It's a little bit different than the information that you get from the proteins, but if we correlate the proteins with the known clinical features we find that the protein modifications are more powerful in this study.JW: In this most recent study, Karin and her team studied nearly 150 uterine tissue samples. And they did so by using a tool called a mass spectrometer. This is an incredibly sensitive instrument that can measure the smallest parts of a sample. Us

Intro: Welcome. I’m your host, Jess Wisse. On today’s episode we’ll talking about how something that may surprise you. Here are a few hints: the holidays, online shopping, and parking. Stay tuned to learn more.   JW: If you’re anything like me, I’m sure you’ve been busy preparing for the holidays. This includes wrapping gifts, RSVPing to parties, and online shopping. LOTS of online shopping. But have you ever considered all of the hands who’ve touched your latest shipment before it arrives on your doorstep? PNNL user experience scientist, Lyndsey Franklin thinks about this daily.LF: Maybe I should set out, you know, Gatorade and some snack bars. No, forget giving milk and cookies to Santa. Give Gatorade and snack bars to your poor delivery drivers because they are hustling. It's insane!JW: Researchers at Pacific Northwest National Laboratory are using their expertise in artificial intelligence, machine learning, and app development to ease challenges with urban freight delivery, an experience especially difficult during the holidays.Meet one of the researchers at PNNL working on this. LF: I am a User Experience Research Scientist in the Visual Analytics Group at PNNL. I try to make computers better playmates for people. I kind of take the philosophy that if something goes wrong, it's the computers fault. It wasn't designed well. Or people didn't think about it well, and it's really not the person's fault. Somebody needs to make the machine play better.JW: Lyndsey is working on a project that’s funded by DOE’s Office of Energy Efficiency and Renewable Energy’s Vehicle Technologies Office. The project is led by the University of Washington’s Urban Freight Lab. Lyndsey’s goal: to develop, test, and improve technologies aimed at cutting time spent by delivery drivers at the curb. Think of your average UPS or FedEx delivery driver. Lyndsey’s job is to increase their productivity and reduce the time and fuel spent searching for available parking.LF: What if you could make something like parking in downtown Seattle smarter? The particular problem that they were trying to address was, ‘How do we make that delivery process more fuel efficient in crazy environments like Seattle?’ So, they came to us in two capacities. They were looking for some expertise in the modeling aspect. John Feo here at PNNL is leading up that part of the of the effort. The other thing that was important to them, and very insightful of them, was to realize that this wasn't going to be something they were designing for a typical desktop environment. This this wasn't build a model, have it run on a big heavy machine, and spit out an answer, and give it to drivers. Downtown Seattle and the parking in downtown Seattle is ever-changing and always chaotic. And so, there's this noisy busy environment and you're supposed to be giving information to drivers who aren't sitting at a computer. So, what is that whole experience that the drivers are going to have. What's that going to look like?JW: Finding parking can be a major headache for freight delivery drivers. Especially in cities like Seattle. Restaurants need a constant cycle of fresh produce. Retail stores depend on delivered products to maintain a steady flow of sales. People living in a city’s apartment buildings expect their Amazon purchases delivered on time, without fail. That’s why PNNL is working with the University of Washington to develop an app that helps drivers identify open parking closest to a delivery location. The Urban Freight Lab calls this sweet spot for a delivery the “final 50 feet”—where a delivery driver stops to deliver their freight.LF: So, the focus of the app is trying to help increase awareness for when parking might be available. In the case of newer drivers who maybe are seasonal, they've been added to routes to deliver Christmas packages and things like that. They're not as familiar with the area. They don't really have that internal map in their head of: ‘Well, if I can't

Pods of Science | Episode 4 | How to Predict Your Next Doctor’s Appointment Intro:Welcome. I’m your host, Jess Wisse. On today’s episode we’ll talking about how artificial intelligence could take your doctor’s care to the next level. Stay tuned to learn more.   MusicJW: PNNL scientists have found a way to improve the accuracy of patient diagnosis by up to 20 percent! How? By using artificial intelligence. A PNNL project, called DeepCare, looked at ways to use AI to improve medical outcomes for patients. Meet the project lead, Robert Rallo. RR: I joined the lab three years ago, coming from Barcelona. My background is chemistry, but I was a professor in computer science for more than 20 years before joining the lab. My main area of expertise is machine learning and applications of machine learning in different areas, one of them being computational toxicology. The team working on this is different computer science scientists from PNNL, Khushbu Agarwal and Sutanay Choudhury. They are two computer scientists in the data sciences group at PNNL. We have strong collaborations also with the University Virginia Tech and Stanford and some of the students who have been summer students here at PNNL have been involved also in this type of biomedical work.JW: We asked Robert why he got into the field of computer science. And here’s what he had to say:RR: The fact that a computer is able to learn by itself from data is something that is really interesting for me, really intriguing, and what triggered my interest in this for me.JW: Robert and his team at PNNL created a new embedding approach. The approach seeks to capture and re-create the types of connections physicians do naturally, in their heads, when they apply a lifetime of learning and knowledge to the patient standing before them in the exam room.What’s embedding? Basically, it’s translation for computers. Using embeddings, computer scientists can take a piece of information that only humans can understand and then transform it into something a computer can use. RR: Medical concepts is, for instance, when you have a specific diagnose this is a concept. You have fever, you have high blood pressure; these are concepts. And then, the way in which a machine learning algorithm or a computer can process these concepts requires them to be codified in a certain numerical way. So one of the ways in which we are making this coding is by developing a continuous numeric representation of these concepts that somehow captures the similarities, the relationships between each one of these individual concepts. So this idea of somehow transforming this textual set of concepts information into a representation which is suitable for machine learning is the embedding process. And what we want is that, this numeric representation will convey the same semantics, the same information, than the original concepts.JW: One of the hardest parts about using AI in the medical field is the inability to combine multiple types of data. Think of all the information that’s captured when you go to the doctor. Now think of all the different forms it comes in. Computer-friendly data like blood work numbers or diagnosis codes are easier than unstructured data like chart notes or images from X-rays and MRIs.RR: Well everybody knows that it’s a known fact that understanding hand-written doctors’ notes is like impossible. (laughing) And I say this because my sister is a medical doctor. But no, I'm joking now. But essentially if we are looking at different types of information, you have structural information in which everything is well classified, well cataloged, and it’s very easy to use. And then you have all this unstructured information in which you have maybe recordings of the patients in an interview for something related to mental health. You can have the notesof doctor that can be written in different narrative styles. You can have different types of imaging data from x-rays to MRI. And each one of these mo

Pods of Science | Episode 3 | How to Do Science Faster with Artificial Intelligence

Intro:

Welcome. I’m your host, Jess Wisse. On today’s episode we’ll talking about a new research center created by the U.S. Department of Energy. Stay tuned to learn more.

Music

JW: Pacific Northwest National Laboratory, Sandia National Laboratories. Georgia Institute of Technology. What do these have in common? They are three powerhouses in the realm of artificial intelligence, and now they are working together.

Want to know who the man is at the helm of this new collaboration? Meet, Roberto.

RG: My name is Roberto Gioiosa. I'm a senior computer scientist at PNNL in the high-performance computing group. My background is in hardware and software core design and mostly focus on the design of operating systems, runtimes, and programming model in particularly looking at emerging a future architecture both for processing, memory, and networking.

I came to PNNL in 2012 after years of other experiences in both academia and industry. Ever since I joined PNNL, I've been trying to lead efforts on software and novel hardware for computational scientists to speed up the solution to their problem and therefore solve scientific challenges.

JW: Artificial intelligence and machine learning seems to be cropping up everywhere these days. From self driving cars to your new smart phone, its everywhere. Even Alexa, Amazon’s voice assistant, is getting smarter with each passing day. Soon, she will be able to guess what you might be thinking with a new feature called Alexa Hunches.

Originally called thinking machines in the 1950s, artificial intelligence is a sub-field of computer science where machines develop the ability to think and learn on their own. Artificial intelligence, also known as AI, allows computers to perform tasks that historically could only be done by humans; think of things like visual perception, speech recognition, language translation. And that’s just the beginning.

RG: Artificial intelligence and machine learning is something that is helping us revolutionize the way we do research. Rather than starting from the top and using first principle to solve a problem, we are trying to see what the data tells us about the problem. This is something that you see every day— look at natural phenomenon and you're trying to find a correlation between what you observe and what are the reasons for that the causal relationships that are in there. In some cases, you know this naturally is complicated and it's not easy to go and have a complete understanding of what is happening without knowing anything about the entire process. What AI is doing for us is helping us do reverse engineering of natural phenomenon.

You have probably seen tons of movies about AI and how that can help, but the fundamental thing is we are looking at the data and we are trying to infer the structure of the phenomenon from the data.

JW: Roberto is the director of a new co-design center, known as the Center for Artificial Intelligence-focused Architectures and Algorithms, or (ARIAA). ARIAA is taking AI & machine learning to the next level.

RG: ARIAA is essentially a tool, a means, in which we are trying to understand what are the

requirements from our application domains. In this case, are power grid, cybersecurity, graph analytics, and chemistry, and how artificial intelligence and machine learning can support these domains to allow novel discoveries.

JW: AARIA will explore how AI and machine learning can support four areas that touch virtually every American’s life. Whether we’re aware of it or not we encounter power grid, cybersecurity, graph analytics, and computational chemistry almost every day. These are the disciplines where new medicines are created, where the fate of our online identity lies, it’s how masses of information is analyzed, and where our lights magically turn on with a flip of the switch.

RG: AI is revolutionizing our world. You see that from your mobil

Jess Wisse (JW): What's behind the science and inventions that impact our daily lives? Pacific Northwest National Laboratory’s Pods of Science are the stories of what happens before the breakthrough. Before a technology becomes a house-hold name, before the life-saving drug his pharmacy shelves, before the paper's published - see what happens when great minds meet great challenges.

Music

Welcome. I’m your host, Jess Wisse. On today’s episode we’ll talking about new technology that may give dogs a run for their money. Wondering what we’re talking about? Stay tuned to learn more.

Music

JW: Floppy-eared sniffer dogs made the news earlier this year as the latest change to security procedures in airports. But in the not-too-distant future, they could be replaced. The replacement is a technology that doesn’t require a scratch behind the ear and a treat to its job.

Meet the man behind the tech: Robert Ewing.

Robert Ewing (RE): I like discovering things. I like solving puzzles. I like doing things that I don't think are possible, or challenging. And learning.

JW: Robert Ewing is a scientist at PNNL. And he’s made a discovery that could potentially save lives.

RE: I'm Robert Ewing. I'm a research chemist at the Pacific Northwest National Lab. I've been here for about 13 years. I've studied various analytical techniques for detecting trace of stances explosives and drugs, or some of those compounds. Ionization chemistry as a part of that. The instrumentation that goes along with that. Those are some of the things that I do for fun.

JW: So, what did Robert discover?

He and his team at PNNL developed a technology that’s ultrasensitive. It detects explosive vapors, deadly chemicals, and drugs like methamphetamine and fentanyl with unparalleled accuracy. And it works in seconds.

RE: The technology really stems from using the detector of a mass spectrometer. And that's a way to look at different molecules, to understand what mass is there. And from that you can sort of determine the analyte. Here you're looking at what we did, or one of the challenges, was is the ionization. So, for the mass spec to see a molecule you've got to put a charge on there so it can manipulate that charge, create an electric field, and separate it. And so the ionization process is a way of getting that charge, that electrical charge, onto an individual molecule. And that's really where I've spent a lot of my time—understanding the chemistry around how that ionization process works and how to improve upon it.

With the commercial mass specs that are out there that are pretty sensitive (parts per billion range and stuff) work pretty well. What we did is, we discovered that if you increase the amount of time that the ionization process can occur you can increase the sensitivity. And so the mass spec has a pinhole bringing the ions in from outside. What we do is, we took the ionization source and moved that away from the mass spec, and instead of having a few milliseconds of reaction time we give it two or three seconds and that gave us several orders of magnitude increase in sensitivity.

JW: This technology could be a game-changer for transportation hubs, mail facilities, and other safety and security screening applications, like the ones you see in airports.

Thanks to Robert’s tireless efforts, the system can detect a whole slew of things. Including explosive vapors, like TNT, toxic chemicals similar to nerve agents, and even illicit drugs, like fentanyl, methamphetamine, and cocaine.

The most surprising part? Robert thought this was a problem that was un-solvable. But he kept mulling on the idea, and that got him to think about none other than man’s best friend.

RE: Probably one of the driving force—the ah-hah moments—I always thought that explosive vapor was a challenge that we probably wouldn't overcome. And yet, dogs go out and sniff explosives all the time. Well, I've always wondered what dogs really smell? You know, are they smelling the explosive? A

Jess Wisse: What's behind the science and inventions that impact our daily lives? Pacific Northwest National Laboratories Pods of Science are the stories of what happens before the breakthrough. Before a technology becomes a house-hold name, before the life-saving drug his pharmacy shelves, before the paper's published.

See what happens when great minds meet great challenges.

Music

Welcome, I'm your host Jess Wisse. On today's episode we'll be unveiling new research coming out of PNNL’s Data Sciences and Analytics Group, but before we unpack that here's a bit more information from our co-host Jessica Bernsen.

Music

Jessica Bernsen: Humans are social animals. Nowhere is that more apparent in today's modern world than on social media. Log in to your favorite social media platforms and you'll find a slew of conversations, debates, news, and more. PNNL researchers took all of this data and built a quantitative framework to better understand communication patterns and how information spreads online.

Meet Svitlana.

Svitlana Volkova: My name is Svitlana Volkova and I am a senior scientist at PNNL. I've been here for three years, and my work involves machine learning, deep learning, natural language processing, and computational social science. I work a lot with social data.

Jess Wisse: Findings gathered by Svitlana and her colleagues Maria Glensky and Emily Saldana shed led light on to how cryptocurrency discussions spread and they also could inform artificial intelligence applications used to forecast things like cryptocurrency prices.

Svitlana Volkova: So in this paper we analyzed almost three years worth of data; a lot of discussions, millions of posts, and comments and that’s what makes this research interesting that like we have access to this vast amount of data that we have the techniques and methodology to analyze really fast and draw some insights and scientific conclusions from this data that can in turn inform machine learning and deep learning models to predict the future.

Jessica Bernsen: Nobody really looked into how information about cryptocurrency spreads on reddit specifically and that's what Svitlana and her team did.

Svitlana Volkova: The current data set included the historical rise of the Bitcoin price and we specifically

wanted to look into social signals around this historical event when the price is increasing and then decreasing we wanted to see how social environments are reflecting this change.

We found that across of three coins, the discussion spread is very different.

We know that Bitcoin is the most popular coin, and that was reflected in our analysis.

We found that comments on a Bitcoin post about was the fastest—on average people responded in 11 minutes to discussions about Bitcoin versus Monero and Ethereum.

In Ethereum threads, it takes people at least 30 minutes to follow up on a on a post, but interestingly we found that Monero has really long, ongoing conversations compared to Bitcoin conversations that have a very short life time.

They don’t live long.

And Bitcoin conversation focus on a specific audience, which on average is between 2 and 6 people. Monero conversations involve more people, and more diverse audiences. And structurally the discussions are very different.

The Monero discussions are like chains. They go deep. And at each level they have a specific size of the audience. Bitcoin discussions are more diverse, and they form trees, and they go more viral compared to Monero.

Jess Wisse: Svitlana and her team looked in to reddit, but their research can also be applied to a variety of platforms. For example, the framework they designed for measuring information spread can be also be extended to measure the spread of other types of information. Such as images on Instagram, videos on YouTube, hashtags on Twitter.

Svitlana Volkova: This analysis would be very helpful for a different predictive analytics so for example you can look how discussions spread around different cryptocu