- Computer searches telescope data for evidence of distant planets
- Sprouting greenery and community
- Scientists find different cell types contain the same enzyme ratios
Posted: 29 Mar 2018 08:59 PM PDT
As part of an effort to identify distant planets hospitable to life, NASA has established a crowdsourcing project in which volunteers search telescopic images for evidence of debris disks around stars, which are good indicators of exoplanets.
Using the results of that project, researchers at MIT have now trained a machine-learning system to search for debris disks itself. The scale of the search demands automation: There are nearly 750 million possible light sources in the data accumulated through NASA's Wide-Field Infrared Survey Explorer (WISE) mission alone.
In tests, the machine-learning system agreed with human identifications of debris disks 97 percent of the time. The researchers also trained their system to rate debris disks according to their likelihood of containing detectable exoplanets. In a paper describing the new work in the journal Astronomy and Computing, the MIT researchers report that their system identified 367 previously unexamined celestial objects as particularly promising candidates for further study.
The work represents an unusual approach to machine learning, which has been championed by one of the paper's coauthors, Victor Pankratius, a principal research scientist at MIT's Haystack Observatory. Typically, a machine-learning system will comb through a wealth of training data, looking for consistent correlations between features of the data and some label applied by a human analyst — in this case, stars circled by debris disks.
But Pankratius argues that in the sciences, machine-learning systems would be more useful if they explicitly incorporated a little bit of scientific understanding, to help guide their searches for correlations or identify deviations from the norm that could be of scientific interest.
"The main vision is to go beyond what A.I. is focusing on today," Pankratius says. "Today, we're collecting data, and we're trying to find features in the data. You end up with billions and billions of features. So what are you doing with them? What you want to know as a scientist is not that the computer tells you that certain pixels are certain features. You want to know 'Oh, this is a physically relevant thing, and here are the physics parameters of the thing.'"
The new paper grew out of an MIT seminar that Pankratius co-taught with Sara Seager, the Class of 1941 Professor of Earth, Atmospheric, and Planetary Sciences, who is well-known for her exoplanet research. The seminar, Astroinformatics for Exoplanets, introduced students to data science techniques that could be useful for interpreting the flood of data generated by new astronomical instruments. After mastering the techniques, the students were asked to apply them to outstanding astronomical questions.
For her final project, Tam Nguyen, a graduate student in aeronautics and astronautics, chose the problem of training a machine-learning system to identify debris disks, and the new paper is an outgrowth of that work. Nguyen is first author on the paper, and she's joined by Seager, Pankratius, and Laura Eckman, an undergraduate majoring in electrical engineering and computer science.
From the NASA crowdsourcing project, the researchers had the celestial coordinates of the light sources that human volunteers had identified as featuring debris disks. The disks are recognizable as ellipses of light with slightly brighter ellipses at their centers. The researchers also used the raw astronomical data generated by the WISE mission.
To prepare the data for the machine-learning system, Nguyen carved it up into small chunks, then used standard signal-processing techniques to filter out artifacts caused by the imaging instruments or by ambient light. Next, she identified those chunks with light sources at their centers, and used existing image-segmentation algorithms to remove any additional sources of light. These types of procedures are typical in any computer-vision machine-learning project.
But Nguyen used basic principles of physics to prune the data further. For one thing, she looked at the variation in the intensity of the light emitted by the light sources across four different frequency bands. She also used standard metrics to evaluate the position, symmetry, and scale of the light sources, establishing thresholds for inclusion in her data set.
In addition to the tagged debris disks from NASA's crowdsourcing project, the researchers also had a short list of stars that astronomers had identified as probably hosting exoplanets. From that information, their system also inferred characteristics of debris disks that were correlated with the presence of exoplanets, to select the 367 candidates for further study.
"Given the scalability challenges with big data, leveraging crowdsourcing and citizen science to develop training data sets for machine-learning classifiers for astronomical observations and associated objects is an innovative way to address challenges not only in astronomy but also several different data-intensive science areas," says Dan Crichton, who leads the Center for Data Science and Technology at NAASA's Jet Propulsion Laboratory. "The use of the computer-aided discovery pipeline described to automate the extraction, classification, and validation process is going to be helpful for systematizing how these capabilities can be brought together. The paper does a nice job of discussing the effectiveness of this approach as applied to debris disk candidates. The lessons learned are going to be important for generalizing the techniques to other astronomy and different discipline applications."
"The Disk Detective science team has been working on its own machine-learning project, and now that this paper is out, we're going to have to get together and compare notes," says Marc Kuchner, a senior astrophysicist at NASA's Goddard Space Flight Center and leader of the crowdsourcing disk-detection project known as Disk Detective. "I'm really glad that Nguyen is looking into this because I really think that this kind of machine-human cooperation is going to be crucial for analyzing the big data sets of the future."
Posted: 29 Mar 2018 12:00 PM PDT
MacGregor House, situated between the Charles River and the MIT soccer fields, is a dorm known for its strong sense of community. Suites of the dormitory (called "entries") each have their own personalities, and the people who live within them often consider each other family. Part of their togetherness stems from the fact that MacGregor is a cook-for-yourself dorm. Students share communal kitchen space and come together to cook and eat dinner most nights of the week.
Even though residents enjoy cooking for themselves, when the academic and extracurricular commitments of a full MIT semester kick-in, free time to shop for fresh produce decreases.
Rachel Weissman, a first-year student studying urban studies and planning at MIT, heads a project to grow fresh produce in the halls of MacGregor. With support from the MindHandHeart Innovation Fund, a grant program promoting wellness, community, and inclusion on campus, the project set up a seven-box hydroponic garden inside the dorm in January. This spring, MacGregor residents will be able to pick tomatoes, bib lettuce, spinach, parsley, and basil from the boxes.
In addition to providing fresh vegetables, the project offers a healthy stress-reliever for the more than 15 students involved in its creation. "There's a component of gardening which is relaxing on its own," says Weissman. "We wouldn't have been able to do it without the MindHandHeart Innovation Fund."
The group hopes to expand the garden, and the participation among residents of MacGregor. "I'd just like to give a big thank you to the heads of house," Weissman says. "We're glad to have this opportunity, and we're excited to see where it goes as more people get involved in the future."
Posted: 29 Mar 2018 09:00 AM PDT
By studying bacteria and yeast, researchers at MIT have discovered that vastly different types of cells still share fundamental similarities, conserved across species and refined over time. More specifically, these cells contain the same proportion of specialized proteins, known as enzymes, which coordinate chemical reactions within the cell.
To grow and divide, cells rely on a unique mixture of enzymes that perform millions of chemical reactions per second. Many enzymes, working in relay, perform a linked series of chemical reactions called a "pathway," where the products of one chemical reaction are the starting materials for the next. By making many incremental changes to molecules, enzymes in a pathway perform vital functions such as turning nutrients into energy or duplicating DNA.
For decades, scientists wondered whether the relative amounts of enzymes in a pathway were tightly controlled in order to better coordinate their chemical reactions. Now, researchers have demonstrated that cells not only produce precise amounts of enzymes, but that evolutionary pressure selects for a preferred ratio of enzymes. In this way, enzymes behave like ingredients of a cake that must be combined in the correct proportions and all life may share the same enzyme recipe.
"We still don't know why this combination of enzymes is ideal," says Gene-Wei Li, assistant professor of biology at MIT, "but this question opens up an entirely new field of biology that we're calling systems level optimization of pathways. In this discipline, researchers would study how different enzymes and pathways behave within the complex environment of the cell."
Li is the senior author of the study, which appears online in the journal Cell on March 29, and in print on April 19. The paper's lead author, Jean-Benoît Lalanne, is a graduate student in the MIT Department of Physics.
An unexpected observation
For more than 100 years, biologists have studied enzymes by watching them catalyze chemical reactions in test tubes, and — more recently — using X-rays to observe their molecular structure.
And yet, despite years of work describing individual proteins in great detail, scientists still don't understand many of the basic properties of enzymes within the cell. For example, it is not yet possible to predict the optimal amount of enzyme a cell should make to maximize its chance of survival.
The calculation is tricky because the answer depends not only on the specific function of the enzyme, but also how its actions may have a ripple effect on other chemical reactions and enzymes within the cell.
"Even if we know exactly what an enzyme does," Li says, "we still don't have a sense for how much of that protein the cell will make. Thinking about biochemical pathways is even more complicated. If we gave biochemists three enzymes in a pathway that, for example, break down sugar into energy, they would probably not know how to mix the proteins at the proper ratios to optimize the reaction."
The study of the relative amounts of substances — including proteins — is known as "stoichiometry." To investigate the stoichiometry of enzymes in different types of cells, Li and his colleagues analyzed three different species of bacteria — Escherichia coli, Bacillus subtilis, and Vibrio natriegens — as well as the budding yeast Saccharomyces cerevisiae. Among these cells, scientists compared the amount of enzymes in 21 pathways responsible for a variety of tasks including repairing DNA, constructing fatty acids, and converting sugar to energy. Because these species of yeast and bacteria have evolved to live in different environments and have different cellular structures, such as the presence or lack of a nucleus, researchers were surprised to find that all four cells types had nearly identical enzyme stoichiometry in all pathways examined.
Li's team followed up their unexpected results by detailing how bacteria achieve a consistent enzyme stoichiometry. Cells control enzyme production by regulating two processes. The first, transcription, converts the information contained in a strand of DNA into many copies of messenger RNA (mRNA). The second, translation, occurs as ribosomes decode the mRNAs to construct proteins. By analyzing transcription across all three bacterial species, Li's team discovered that the different bacteria produced varying amounts of mRNA encoding for enzymes in a pathway.
Different amounts of mRNA theoretically lead to differences in protein production, but the researchers found instead that the cells adjusted their rates of translation to compensate for changes in transcription. Cells that produced more mRNA slowed their rates of protein synthesis, while cells that produced less mRNA increased the speed of protein synthesis. Thanks to this compensation, the stoichiometry of enzymes remained constant across the different bacteria.
"It is remarkable that E. coli and B. subtilis need the same relative amount of the corresponding proteins, as seen by the compensatory variations in transcription and translation efficiencies," says Johan Elf, professor of physical biology at Uppsala University in Sweden. "These results raise interesting questions about how enzyme production in different cells have evolved."
"Examining bacterial gene clusters was really striking," lead author Lalanne says. "Over a long evolutionary history, these genes have shifted positions, mutated into different sequences, and been bombarded by mobile pieces of DNA that randomly insert themselves into the genome. Despite all this, the bacteria have compensated for these changes by adjusting translation to maintain the stoichiometry of their enzymes. This suggests that evolutionary forces, which we don't yet understand, have shaped cells to have the same enzyme stoichiometry."
In the future, Li and his colleagues will test whether their findings in bacteria and yeast extend to humans. Because unicellular and multicellular organisms manage energy and nutrients differently, and experience different selection pressures, researchers are not sure what they will discover.
"Perhaps there will be enzymes whose stoichiometry varies, and a smaller subset of enzymes whose levels are more conserved," Li says. "This would indicate that the human body is sensitive to changes in specific enzymes that could make good drug targets. But we won't know until we look."
Beyond the human body, Li and his team believe that it is possible to find simplicity underlying the complex bustle of molecules within all cells. Like other mathematical patterns in nature, such as the the spiral of seashells or the branching pattern of trees, the stoichiometry of enzymes may be a widespread design principle of life.
The research was funded by the National Institutes of Health, Pew Biomedical Scholars Program, Sloan Research Fellowship, Searle Scholars Program, National Sciences and Engineering Research Council of Canada, Howard Hughes Medical Institute, National Science Foundation, Helen Hay Whitney Foundation, Jane Coffin Childs Memorial Fund, and the Smith Family Foundation.
|You are subscribed to email updates from MIT News. |
To stop receiving these emails, you may unsubscribe now.
|Email delivery powered by Google|
|Google, 1600 Amphitheatre Parkway, Mountain View, CA 94043, United States|