Subterranean Pharmacy, developments in 6G and an AI that can interpret Brain Scans.
March 23
This week we will investigate a Subterranean Pharmacy that may provide us with a new source of antibiotics. We also examine a faster camera that allows us to investigate the atomic scale of matter and we look at an advance in the equipment required for 6G mobile phones. Finally we meet an AI that can interpret brain scans and tell us what we have been looking at.
Subterranean Pharmacy
A team from the University of Liege in Belgium have found that moonmilk, a mineral deposit found in caves, contains a cryptic compound that is active against bacteria that are resistant to multiple antibiotics. Moonmilk is found in various forms (pasty, dry or liquid) in limestone caves. There is archeological evidence that moonmilk was used as an anti-infectious agent many years ago by people living in the Swiss and Austrian Alps.
The team had been studying the microbial flora of moonmilk in order to find out if it had a use in human and animal therapy. They visited the caves of the Condruzian plateau to find deposits of moonmilk and to isolate filamentous bacteria. The team found a wide range of bacteria both in number and diversity.
Analyses of the genomes of the bacteria retrieved from the caves revealed that they participate together in what is a vast subterranean pharmacy. The consortium of microbes is capable of producing hundreds of antibiotics, some of which we already use daily.
The most unusual finding was that the majority of the biosynthesis genes involved in producing the bioactive compounds were cryptic. That is, it is not possible to associate them with a known molecule. This means that these strains constitute a real reservoir for the discovery of new bioactive molecules.
Unfortunately the molecules do not lend themselves to be easily grown under laboratory conditions. After they have been extracted from the extreme conditions of the caves it is difficult to keep them ‘alive’ once they have been brought to the surface.
A startup called, HEDERA-22 has been established to continue the research and commercialize the findings. The first cryptic compound that they have figured out is an antibiotic called lunaemycin which is produced by a new bacterium called Streptomycin lunaelactis. It is potentially useful against Gram-positive bacteria that are multi resistant to antibiotics.
It is not yet known if this antibiotic will be suitable for human use however it is a proof of concept that new potentially useful molecules can be discovered in subterranean caves.
Faster Camera Shutter
Researchers at Columbia Engineering and Universite de Bourgogne (in Dijon, France) have developed a new type of camera that has a variable shutter speed of around 1 trillionth of a second. This new camera allows them to see the dynamic disorder of atoms.
It appears that the best performing materials in sustainable energy applications (e.g. conversion of sunlight or waste heat to electricity) use the collective fluctuations of clusters of atoms within a much larger structure. This process is knows as dynamic disorder. When materials function within an operating device they can behave as if they are alive. Parts of the material respond and change in unexpected ways. They are difficult to study as the clusters are so small and they fluctuate in time.
When looking at dynamic disorder with a slow shutter speed the dynamic disorder is blurred however with the new camera they could see the disorder. This visualization shows the difference.
The new camera is not like a conventional camera. It uses neutrons to measure atomic positions with a shutter speed of a picosecond. The visualization allows the researchers to untangle the complexity of what is going on in complex materials. They can watch a material and see which atoms are in the dance and which are sitting it out (so wallflowers also exist at the atomic level).
Further development of the system should allow it to be able to measure any material where atomic dynamics are important. This includes watching lithium moving around battery electrodes and to study the dynamic process of water splitting with sunlight. The team has already observed cubic GeTe atomic symmetries being broken. GeTe is an important material in waste heat to electricity conversion. This breakthrough will allow researchers to look for new materials with the observed effects leading to even better materials.
Tunable meta-devices for 6G Communications
We have spoken previously about the projected delivery of 6G phone systems by around 2030 (Samsung has said 2028). To achieve this goal many new technologies need to be developed. 6G takes advantage of the terahertz part of the spectrum. We need specific new technology to work at these higher speeds.
A team at City University in Hong Kong have invented a tunable terahertz meta device that can control radiation direction and coverage of terahertz beams. By rotating its meta surface the device can direct the 6G signal to a designated recipient (and only that recipient). This minimizes power leakage and enhances privacy.
Current terahertz systems are bulky, use heavy lenses and reflectors which can only guide waves to fixed transmitters or detectors or transmit them to a single receiver at a fixed location in a limited coverage area. 6G applications require precise positioning and concentrated signal strength.
The device consists of several artificial thin sheets with sub wavelength thickness. These surfaces work as efficient projects to steer the focal point of THz beams. The 30mm surface has about 11,000 micro antennas. By rotating the surfaces the THz beam can be directed to a specified (via X, Y and Z coordinates) location.
The meta surfaces are fabricated with a high temperature resin and a 3D printing method. They are lightweight, small and can be easily produced at large scale at low cost. 6G devices should be able to support 100 gigabits per second and potentially terabits (1000 gigabits). That is hundreds to thousands of times faster than the 5G transmission rate.
AI recreates what we see from Brain Scans
Neuroscientists have difficulty in demystifying how the human brain converts what our eyes see, into mental images. AI however is getting much better at mimicking the brain. A team from Osaka University have demonstrated an AI that can read brain scans and recreate largely realistic versions of the image that a person has seen.
The team used the Stable Diffusion algorithm to recreate the images. Stable Diffusion is one of several text to image AI’s released into the market in 2022. The team added additional training to the standard Stable Diffusion system, linking additional text descriptions about thousands of photos to brain patterns that were exhibited by people viewing the photos.
The AI was able to use information gathered from different regions of the brain including the occipital and temporal lobes. The system interpreted information from fMRI brain scans which detect changes in blood flows to active regions of the brain. When people look at a photo the temporal lobes register information about contents of the image (people, objects or scenery) and the occipital lobe registers information about layout and perspective. The MRI captures peaks in brain activity and these patterns can then be reconverted into an imitation image using the AI.
Currently the system is limited to using the scans of people who provided the training brain scans. Adding more individuals would require more training of the AI however the team is hoping that the technology can be developed to image thoughts and dreams. It could also be used to understand how different animals perceive reality.
Paying it Forward
If you have a start-up or know of a start-up that has a product ready for market please let me know. I would be happy to have a look and feature the startup in this newsletter. Also if any startups need introductions please get in touch and I will help where I can.
If you have any questions or comments please comment below.
I would also appreciate it if you could forward this newsletter to anyone that you think might be interested.
Till next week.