Connect with us

Tech

Can all of your digital photographs be stored as DNA?

Published

on

DNA Data Storage

[ad_1]

DNA data storage

Biological engineers at MIT have demonstrated a way to easily retrieve data files stored as DNA. This could be a step towards using DNA archives to store vast amounts of photographs, images, and other digital content. Credit: Image: MIT News. Small icons courtesy of the researchers

A method of labeling and extracting DNA data files from a large pool can help make DNA data storage possible.

There are now about 10 trillion gigabytes of digital data on Earth, and every day people create emails, photos, tweets and other digital files that add up to another 2.5 million gigabytes of data. Much of this data is stored in huge facilities known as exabyte data centers (exabyte is 1 billion gigabytes), which can be the size of several football fields and cost about $ 1 billion to build and maintain.

Many scientists believe that the alternative solution lies in a molecule that contains our genetic information: DNA, which has evolved to store vast amounts of information at very high densities. According to Mark Bath, professor of biological engineering at MIT, a coffee mug full of DNA could theoretically store all the world’s data.

“We need new solutions to store the huge amounts of data that is accumulating in the world, especially archival data,” says Bath, who is also an associate member of the Broad Institute at MIT and Harvard. “DNA is a thousand times denser than even flash memory, and another interesting property is that once you create a DNA polymer, it doesn’t consume energy. You can write down DNA and keep it forever. “

DNA files photo

Photo of DNA “files”. Each silica sphere contains DNA sequences that code for a particular image, and the outside of the sphere is covered with nucleotide barcodes that describe the contents of the image. Credit: Courtesy of the researchers

Scientists have already demonstrated that they can encode images and pages of text like DNA. However, you also need an easy way to select the file you want from a variety of DNA fragments. Batch and his colleagues demonstrated one way to do this by encapsulating each data file in a 6-micrometer particle of silica, which is labeled with short DNA sequences that reveal the contents.

Using this approach, the researchers demonstrated that they can accurately extract individual images stored as DNA sequences from a set of 20 images. Given the number of possible labels that can be used, this approach can be scaled up to 10twenty files.

Bath is the senior author of the study, which is published today in Materials of Nature… The lead authors of the article are MIT Senior Postdoc James Banal, former MIT Research Fellow Tyson Shepherd, and MIT Graduate Student Joseph Berleant.

Stable storage

Digital storage systems encode text, photographs, or any other information as a sequence of ones and zeros. This same information can be encoded in DNA using the four nucleotides that make up the genetic code: A, T, G, and C. For example, G and C can be used to represent 0, while A and T represent 1.

DNA has several other characteristics that make it desirable as an information carrier: it is extremely stable, and it is fairly easy (but expensive) to synthesize and sequenced. Plus, because of its high density – each nucleotide equivalent to two bits is about 1 cubic nanometer – an exabyte of data stored as DNA can fit in the palm of your hand.

Images stored in DNA

Researchers have stored similar images in DNA. Credit: Courtesy of the researchers

One obstacle to this type of data storage is the cost of synthesizing such large amounts of DNA. Currently, it will cost $ 1 trillion to write one petabyte of data (1 million gigabytes). To compete with magnetic tape, which is often used for archival data storage, Bath estimates the cost of DNA synthesis would need to drop by about six orders of magnitude. Bath says he expects this to happen within a decade or two, much like the cost of storing information on flash drives has plummeted over the past couple of decades.

Aside from cost, another major bottleneck when using DNA to store data is the difficulty of choosing the right file from all the others.

“Assuming DNA recording technologies reach the point where it is economically viable to write exabytes or zettabytes of data into DNA, what then? You will have a bunch of DNA, which is millions of files, images, films and so on, and you will need to find the picture or film that you are looking for, ”says Bath. “It’s like trying to find a needle in a haystack.”

Nowadays, DNA files are usually obtained using PCR (polymerase chain reaction). Each DNA data file includes a sequence that binds to a specific PCR primer. To extract a specific file, this primer is added to the sample to find and amplify the desired sequence. However, one of the drawbacks of this approach is that cross-interactions can occur between the primer and DNA sequences outside the target, leading to the removal of unwanted files. In addition, the PCR production process requires enzymes and ultimately consumes most of the DNA that was in the pool.

“You kind of burn a haystack to find the needle, because all the rest of the DNA isn’t amplified, and you’re essentially throwing it away,” says Bath.

Retrieving the file

As an alternative approach, the MIT team has developed a new extraction technique that involves encapsulating each DNA file in a small particle of silica. Each capsule is labeled with single-stranded DNA “barcodes” corresponding to the contents of the file. To demonstrate this approach in a cost-effective way, the researchers encoded 20 different images into DNA fragments of about 3,000 nucleotides in length, which is the equivalent of about 100 bytes. (They also showed that the capsules fit DNA files up to a gigabyte in size.)

Each file was labeled with barcodes corresponding to labels such as “cat” or “plane.” When researchers want to extract a specific image, they remove the DNA sample and add primers that match the tags they are looking for, such as “cat”, “orange” and “wild” for the image. tiger, or “cat”, “orange” and “domestic” for a domestic cat.

The primers are labeled with fluorescent or magnetic particles, making it easy to extract and identify any matches from the sample. This allows the desired file to be removed while leaving the rest of the DNA intact to be returned to storage. Their search process allows for logical statements such as “President AND 18th century ”to get the result of George Washington, similar to what you get when you search for images on Google.

“At the current stage of testing our concept, the search speed is 1 kilobyte per second. The search speed of our file system is determined by the size of data per capsule, which is currently limited by the prohibitive cost of writing even 100 megabytes of data per DNA, and the number of sorters that we can use in parallel. If DNA synthesis becomes cheap enough, we can maximize the amount of data that we can store on file with our approach, ”says Banal.

For their barcodes, the researchers used single-stranded DNA sequences from a library of 100,000 sequences, each about 25 nucleotides long, developed by Stephen Elledge, professor of genetics and medicine at Harvard Medical School. If you put two such labels on each file, you can uniquely label 1010 (10 billion) different files, and with four shortcuts on each you can uniquely tag 10twenty files.

George Church, professor of genetics at Harvard Medical School, describes this technique as “a giant leap forward in knowledge management and search technology.”

“Rapid progress in writing, copying, reading and storing energy-efficient archival data in DNA form has left poorly understood possibilities for accurately extracting data files from huge (1021 years byte, zetta-scale) of databases, ”says Church, who was not involved in the study. “The new research effectively addresses this problem by using a completely independent outer layer of DNA and different properties of DNA (hybridization, not sequencing), and moreover, using existing tools and chemistry.”

Bat suggests that this kind of DNA encapsulation can be useful for storing “cold” data, that is, data that is archived and rarely accessed. His lab is creating a startup Cache DNA, which is now developing technology for long-term DNA storage, both for storing DNA data in the long term and for clinical and other pre-existing DNA samples in the near future.

“While it may take some time before DNA becomes viable as a data carrier, there is already an urgent need for low-cost, massive storage solutions for pre-existing DNA and RNA samples from Covid-19 testing, human genomic sequencing, and more. the field of genomics, ”says Bath.

Reference: “Random Access DNA Memory Using Boolean Search in Archive File Storage” by James L. Banal, Tyson R. Shepherd, Joseph Berleant, Hellen Huang, Miguel Reyes, Cherie M. Ackerman, Paul S. Blainey, and Mark Bath. June 10, 2021, Materials of Nature
DOI: 10.1038 / s41563-021-01021-3

The study was funded by the Office of Naval Research, the National Science Foundation, and the US Army Research Office.



[ad_2]

Originally reported by Source link

The featured images are, as they appear on the original report.

Tech

Researchers reconfigure material topology at microscale

Published

on

[ad_1]

TO

Encrypted Harvard Shield

The researchers encoded the patterns and designs in the material, making tiny, invisible changes to the geometry of the triangular lattice. Credit: Image courtesy of Shukong Lee / Bolei Deng / Harvard SEAS.

Reconfigurable materials can do amazing things. Flat sheets turn to faceExtruded cube transforms into dozens of different shapes… But there is one thing that a reconfigurable material cannot change yet: its underlying topology. A 100 cell reconfigurable material will always have 100 cells, even if those cells are stretched or compressed.

Now researchers at the Harvard School of Engineering and Applied Sciences. John A. Paulson (SEAS) developed a method for altering the fundamental topology of cellular material at the microscale. Research published in Nature

“The creation of cellular structures that can dynamically change their topology will open up new possibilities in the development of active materials with information encryption, selective particle capture, and customizable mechanical, chemical and acoustic properties,” said Joanna Eisenberg, professor of materials science Amy Smith Berilson. at SEAS, professor of chemistry and chemical biology and senior article author.

Triangle material topology

Hexagon material topology

Researchers have developed a method for altering the fundamental topology of a cellular material at the microscale, paving the way for active materials with tunable mechanical, chemical and acoustic properties. Credit: Images courtesy of Shukong Lee / Bolei Deng / Harvard SEAS.

The researchers used the same physics that binds our hair when it gets wet – capillary force. Capillary force works well on soft, pliable material like our hair, but it fights tough cellular structures that require bending, stretching, or folding, especially around strong, connected knots. Capillary force is also temporary as materials tend to return to their original configuration after drying.

To develop a durable but reversible method for transforming the topology of rigid cellular microstructures, the researchers developed a two-tier dynamic strategy. They started with a rigid polymer honeycomb microstructure with a triangular lattice topology and exposed it to droplets of a volatile solvent chosen to swell and soften the polymer at the molecular level. This made the material temporarily more flexible, and in this flexible state, the capillary forces created by the evaporating liquid brought the edges of the triangles closer together, changing their bonds with each other and turning them into hexagons. Then, as the solvent quickly evaporated, the material dried out and was trapped in its new configuration, regaining its stiffness. The whole process took a matter of seconds.

“When you think about applications, it is very important not to lose the mechanical properties of the material after the transformation process,” said Shukong Lee, a graduate student at Eisenberg’s lab and co-author of the paper. “Here we have shown that we can start with a hard material and end with a hard material while temporarily softening it during the reconfiguration phase.”

Assembly of microstructures

Microstructure assembly video. A liquid acts on the triangular lattice, which swells and softens the polymer. In this flexible state, the capillary forces created by the evaporating liquid brought the edges of the triangles closer together, changing their bonds with each other and turning them into hexagons. Credit: Video courtesy of Shukong Lee / Bolei Deng / Harvard SEAS

The new material topology is so strong that it can withstand heat or immersion in some fluids for days without disassembly. Its reliability actually posed a problem for researchers who hoped to make the conversion reversible.

To go back to the original topology, the researchers developed a method for combining two fluids. The former temporarily enlarges the lattice, which causes the adhering walls of the hexagons to flake off and allows the lattice to return to its original triangular structure. The second, less volatile liquid delays the emergence of capillary forces until the first liquid evaporates and the material regains its rigidity. Thus, structures can be repeatedly assembled and disassembled and held in any intermediate configuration.

Disassembly of microstructures

Video of disassembly of microstructures. The first one temporarily swells the lattice, which peels off the adhered walls. The second, less volatile liquid delays the emergence of capillary forces until the first liquid evaporates and the material regains its rigidity. Credit: Video courtesy of Shukong Lee / Bolei Deng / Harvard SEAS

“To expand our approach to arbitrary lattices, it was important to develop a generalized theoretical model that links cell geometry, material stiffness, and capillary forces,” said Bolei Deng, paper co-author and graduate student at Katya Bertholdi’s laboratory, William and Ami Kuan Danoff Professor of Applied Mechanics. in SEAS.

Guided by this model, the researchers demonstrated programmed reversible topological transformations of various lattice geometries and reacting materials, including the transformation of a lattice of circles into squares.

Researchers have studied various applications for research. For example, the team encoded the patterns and designs in the material, making tiny invisible changes to the geometry of the triangular lattice.

“You can imagine that this will be used to encrypt information in the future, because you cannot see the pattern in the material when it is disassembled,” Lee said.

The researchers also demonstrated very local transformation, assembly and disassembly of sections of the lattice using a tiny liquid droplet. This method can be used to adjust the friction and wetting properties of a material, change its acoustic properties and mechanical elasticity, and even trap gas particles and bubbles.

“Our strategy can be applied to a variety of applications,” said Bertholdi, who is also a co-author of the article. “We can apply this method to a variety of materials, including sensitive materials, with different geometries and different scales, even at the nanoscale, where topology plays a key role in the development of tunable photonic metasurfaces. The design space for that is huge. “

Reference: “Fluid-Induced Topological Transformations of Cellular Microstructures” by Shukong Lee, Bolei Deng, Alison Grintal, Alisha Schneider-Yamamura, Jinliang Kang, Reese S. Martens, Katie T. Zhang, Jiang Li, Xicin Yu, Katya Bertholdi, and Joanna. Eisenberg, April 14, 2021, Nature
DOI: 10.1038 / s41586-021-03404-7

This study was co-authored by Alison Grintal, Alissha Schneider-Yamamura, Jinliang Kang, Reese S. Martens, Katie T. Zhang, Jian Li, and Xiqing Yu.

It was supported by the National Science Foundation under the Material Design for Revolution and Design for Our Future (DMREF) Program # DMR-1922321, Harvard University Materials Research and Engineering Center (MRSEC), award # DMR-18 2011754 and the Department of Energy USA (DOE), Office of Science, Basic Energy Sciences (BES) under number DE-SC0005247.



[ad_2]

Originally reported by Source link

The featured images are, as they appear on the original report.

Continue Reading

Tech

Help clean up space junk with the first of its kind Mission Optimizer

Published

on

Help clean up space junk with the first of its kind Mission Optimizer

[ad_1]

ClearSpace-1 takes over Vespa

Illustration ClearSpace-1, a planned mission for active garbage collection. Credit: ClearSpace SA.

Less than a year after the UK Space Agency pledged £ 1 million to combat space debris, Fujitsu UK has successfully combined quantum computing and artificial intelligence to support the transformation of its debris removal system.

Fujitsu Prototype – Created in collaboration with Glasgow University, Amazon Web Services and Astroscale UK will improve mission planning so that a single spacecraft can efficiently choose which pieces of space debris to remove in a single mission, much faster than is currently possible.

Removing space debris is key to sustainability in space, reducing or even preventing the risk of legacy spacecraft colliding with new and existing satellites.

What’s more, supporting debris removal missions with Fujitsu technology will help reduce the risk of catastrophic collisions in orbit, which could lead to thousands of other new debris that pose a real threat to operational satellites in orbit.

By carefully deciding what debris and when to collect, Digital Annealer’s quantum proposal optimizes a mission plan to determine the minimum fuel and minimum time needed to safely return a failed spacecraft or satellites to disposal orbit. Finding the optimal route for collecting space debris will significantly save time and money during the mission planning phase, as well as, as a result, increase commercial profitability.

Space debris objects in low Earth orbit

The debris objects shown in the images are artist’s impression based on actual density data. However, debris objects are shown zoomed in to make them visible at the specified scale. Credit: ESA.

With 2350 inoperative satellites Currently in orbit and more than 28,000 debris tracked by space surveillance networks, this technology will help the UK increase its market share in the space sector. It will also further support the UK government’s commitment to a more sustainable future overall.

Dr. Matteo Ceriotti, professor of space engineering, and graduate student Julia Viavattene, who led the project for the University of Glasgow, said: “The University of Glasgow was involved in this project from the very beginning – developing the trajectory. the models needed to efficiently remove space debris as well as to estimate the cost of travel.

“The University has a long history of experience in the design and optimization of space trajectories, which is why we have strived to be at the forefront of any government initiative to enhance the UK’s image in the space sector. With the help of Fujitsu, AWS and Astroscale UK, we have helped overcome the challenges of removing space debris to make future projects easier. ”

Ellen Devereux, digital firing consultant for Fujitsu UK and Ireland, said: “All space debris presents a potential risk of collision with operating systems that many of us take for granted, from weather forecasts to telecommunications.

“With the support of the UK Space Agency, Astroscale UK, AWS and the University of Glasgow, we have developed a solution to optimize the planning of a maintenance ship’s mission before it goes into space. This means that organizations like Astroscale UK can choose to collect more trash faster than ever before.

“This not only makes the process much more cost-effective for organizations that need to move and dispose of garbage, but it also leverages artificial intelligence and quantum computing.

“Over the past six months, we have learned that this technology has tremendous value for optimization in space; not only for garbage collection, but also for in-orbit maintenance and more. We now understand its potential better, and we can’t wait to see how the technology will be applied during the future mission. ”

Jacob Gere, head of space surveillance and tracking at the UK space agency, said: “Monitoring dangerous space objects is vital to protect the services we all rely on, from communications devices to satellite navigation. This project is one of the first examples of the use of quantum computing and artificial intelligence to solve space debris problems, but it is unlikely to be the last.

“The UK is committed to making space sustainable, and Fujitsu, working with Astroscale UK, the University of Glasgow and AWS, has demonstrated the real value of keeping space uncluttered by making it accessible to future generations.”

Stephen Vox, CTO at Astroscale UK, said: “Finding the optimal mission plan manually is time consuming and difficult. Astroscale UK is leading the next pioneering move under the End-of-Life Services by Astroscale (ELSA) program to remove not only one piece of debris, but multiple objects of debris with a single serving satellite known as ELSA-M, which is essentially a more economical way to remove debris from orbit. By teaming up with Fujitsu, AWS and the University of Glasgow, we hope to further optimize this for future missions. ”

The study was carried out under a grant from the UK Space Agency “Expanding Space Surveillance and Tracking Research”. The project, which was developed over six months in accordance with government digital services guidelines, uses fast technologies based on an artificial neural network (ANN). Trajectory design algorithms developed by the University of Glasgow, together with Fujitsu Digital Annealer and Quantum Inspired Optimization Services, to address some of the major optimization problems associated with Active Debris Removal (ADR) mission planning.

Amazon Web Services provided Cloud and AI and ML tools and services to support the project. The Amazon Sagemaker toolbox has been used to rapidly develop ANNs that accurately predict the cost of orbital movements in a fraction of the time required to fully calculate them. Astroscale UK, the world’s first commercial company to embark on a debris removal demonstration mission from lower Earth orbit, offers an end-use case as a typical user of multi-mission optimization.

Fujitsu, which spearheaded the project, is one of seven UK companies to receive more than £ 1 million in space debris tracking from the UK Space Agency. The UK Space Agency and the Department of Defense have announced the next step in their joint initiative to raise UK awareness of events in space.



[ad_2]

Originally reported by Source link

The featured images are, as they appear on the original report.

Continue Reading

Tech

Scientists use human protein to deliver RNA molecular drugs to cells

Published

on

SEND Technology Restoring Health

[ad_1]

Fully assembled SEND packages

The fully assembled SEND bags are released from the collection cage for gene therapy. Credit: McGovern Institute

A programmable system built from human body components is a step towards safer and more targeted delivery of gene editing and other molecular therapeutics.

Researchers from Massachusetts Institute of TechnologyThe McGovern Institute for Brain Research at MIT, Howard Hughes Medical Institute and the Broad Institute at MIT and Harvard have developed a new way to deliver molecular therapy to cells. A system called SEND can be programmed to encapsulate and deliver various RNA cargo. SEND uses natural proteins in the body that form virus-like particles and bind RNA, and may elicit a smaller immune response than other delivery approaches.

The new delivery platform works effectively in cell models and, with further development, could open up a new class of delivery methods for a wide range of molecular drugs, including gene editing and gene replacement. Existing delivery vehicles for these therapeutic agents may be ineffective and randomly integrate into the genome of cells, and some may stimulate unwanted immune responses. SEND promises to overcome these limitations, which could open up new avenues for the introduction of molecular medicine.


Researchers at MIT, MIT’s McGovern Brain Research Institute, Howard Hughes Medical Institute, and MIT’s Broad Institute and Harvard have developed a new way to deliver molecular therapy to cells. Credit: Created by the McGovern Institute and Opus Design in collaboration with Feng Zhang, Rhiannon Macrae, and the Broad Institute.

“The biomedical community is developing powerful molecular therapeutics, but getting them to cells in an accurate and efficient manner is not an easy task,” said CRISPR pioneer Feng Zhang, senior study author, member of the Broad Institute core institute, researcher at McGovern. Institute, and Professor of Neuroscience James and Patricia Poitras at MIT. “SEND can solve these problems.” Zhang is also a researcher at Howard Hughes Medical Institute and a professor in the Department of Brain, Cognitive Sciences, and Biological Engineering at MIT.

I report The science, the command describes how to SEND (Selective Eendogenous eNcapsidization for cell Delivery) uses molecules produced by human cells. At the center of SEND is a protein called PEG10, which usually binds to its own mRNA and forms a spherical protective capsule around it. In their research, the team developed PEG10 to selectively package and deliver other RNAs. The scientists used SEND to deliver the CRISPR-Cas9 gene editing system to mouse and human cells to edit target genes.

First author Michael Segel, a postdoctoral fellow at Zhang’s lab, and Blake Lash, a second author and graduate student at the lab, said PEG10 is not unique in its ability to carry RNA. “That’s what is so exciting,” Segel said. “This study shows that there are probably other RNA transfer systems in the human body that can also be used for therapeutic purposes. It also raises some really interesting questions about what the natural roles of these proteins might be. “

Inspiration from within

The PEG10 protein exists naturally in the human body and comes from the “retrotransposon” – a virus-like genetic element that was integrated into the genome of human ancestors millions of years ago. Over time, PEG10 was metabolized by the body to become part of the repertoire of proteins essential for life.

Four years ago, researchers showed that another protein derived from the retrotransposon, ARC, forms virus-like structures and is involved in the transfer of RNA between cells. Although these studies have shown that it is possible to create retrotransposon proteins as a delivery platform, scientists have not been able to successfully use these proteins to package and deliver specific RNA shipments into mammalian cells.

SEND Health Recovery Technologies

SEND packages are injected into diseased cells to deliver therapeutic mRNA and restore health. Credit: McGovern Institute

Knowing that some proteins derived from the retrotransposon are capable of binding and packaging molecular loads, Zhang’s team turned to these proteins as possible delivery vehicles. They systematically searched for these proteins in the human genome, looking for those that could form protective capsules. In their initial analysis, the team found 48 human genes encoding proteins that could have this ability. Of these, 19 candidate proteins were present in both mice and humans. In the cell line that the team studied, PEG10 stood out as an efficient shuttle; the cells released significantly more PEG10 particles than any other protein tested. PEG10 particles also predominantly contain their own mRNA, suggesting that PEG10 may be able to package certain RNA molecules.

Development of a modular system

To develop the SEND technology, the team identified molecular sequences or “signals” in PEG10 mRNA that PEG10 recognizes and uses to package its mRNA. The researchers then used these signals to construct both PEG10 and other RNA cargoes so that PEG10 could selectively package those RNAs. The team then adorned the PEG10 capsules with additional proteins called “fusogens” that sit on the surface of cells and help them fuse together.

By creating fusogens on PEG10 capsules, researchers must be able to target the capsule to a specific type of cell, tissue, or organ. As a first step towards this goal, the team used two different fusogens, including one found in the human body, to ensure the delivery of the SEND cargo.

“By mixing and matching the various components in the SEND system, we believe it will provide a modular platform for the development of therapeutic agents for a variety of diseases,” Zhang said.

Progressive gene therapy

SEND is made up of proteins that are naturally produced in the body, which means it cannot trigger an immune response. If demonstrated in further research, the researchers say SEND could open the door to multiple uses of gene therapy with minimal side effects. “SEND technology will augment viral delivery vectors and lipid nanoparticles to further expand the toolbox for gene delivery and therapies into cells,” Lash said.

The team will then test SEND on animals and further develop a system to deliver the cargo to various tissues and cells. They will also continue to explore the natural diversity of these systems in the human body to identify other components that can be added to the SEND platform.

“We are delighted to continue to advance this approach,” Zhang said. “The realization that we can use PEG10 and most likely other proteins to engineer a delivery pathway in the human body to package and deliver new RNA and other potential therapies is a really powerful concept.”

Reference: “The mammalian retrovirus-like protein PEG10 packs its own mRNA and can be pseudotyped to deliver mRNA” Michael Segel, Blake Lash, Jingwei Song, Alim Ladha, Catherine S. Liu, Xin Jin, Sergey L. Mehedov, Rhiannon K. Macrae , Evgeny Kunin and Feng Zhang, August 20, 2021, The science
DOI: 10.1126 / science.abg6155

This work was made possible with the support of the Simons Center for the Social Brain at MIT; Internal Research Program of the National Institutes of Health; National Institutes of Health Grants 1R01-HG009761 and 1DP1-HL141201; Howard Hughes Medical Institute; Open charity; G. Harold and Leila Y. Mathers Charitable Foundation; Edward Mullinkrodt Jr. Foundation; The Poitras Center for Mental Disorders Research at the Massachusetts Institute of Technology; Hock E. Tan and C. Lisa Young, Center for Autism Research, MIT; The Yang Tang Center for Molecular Therapy at the Massachusetts Institute of Technology; Lisa Yang; The Phillips family; R. Metcalfe; and J. and P. Poitras.



[ad_2]

Originally reported by Source link

The featured images are, as they appear on the original report.

Continue Reading

Trending