Lifeboat Foundation: Safeguarding Humanity
Toggle light / dark theme

Blog – Page 5309

The yin-yang codec transcoding algorithm is proposed to improve the practicality and robustness of DNA data storage.

Given these results, YYC offers the opportunity to generate DNA sequences that are highly amenable to both the ‘writing’ (synthesis) and ‘reading’ (sequencing) processes while maintaining a relatively high information density. This is crucially important for improving the practicality and robustness of DNA data storage. The DNA Fountain and YYC algorithms are the only two known coding schemes that combine transcoding rules and screening into a single process to ensure that the generated DNA sequences meet the biochemical constraints. The comparison hereinafter thus focuses on the YYC and DNA Fountain algorithms because of the similarity in their coding strategies.

The robustness of data storage in DNA is primarily affected by errors introduced during ‘writing’ and ‘reading’. There are two main types of errors: random and systematic errors. Random errors are often introduced by synthesis or sequencing errors in a few DNA molecules and can be redressed by mutual correction using an increased sequencing depth. System atic errors refer to mutations observed in all DNA molecules, including insertions, deletions and substitutions, which are introduced during synthesis and PCR amplification (referred to as common errors), or the loss of partial DNA molecules. In contrast to substitutions (single-nucleotide variations, SNVs), insertions and deletions (indels) change the length of the DNA sequence encoding the data and thus introduce challenges regarding the decoding process. In general, it is difficult to correct systematic errors, and thus they will lead to the loss of stored binary information to varying degrees.

To test the robustness baseline of the YYC against systematic errors, we randomly introduced the three most commonly seen errors into the DNA sequences at a average rate ranging from 0.01% to 1% and analysed the corresponding data recovery rate in comparison with the most well-recognized coding scheme (DNA Fountain) without introducing an error correction mechanism. The results show that, in the presence of either indels (Fig. 2a) or SNVs (Fig. 2b), YYC exhibits better data recovery performance in comparison with DNA Fountain, with the data recovery rate remaining fairly steady at a level above 98%. This difference between the DNA Fountain and other algorithms, including YYC, occurs because uncorrectable errors can affect the retrieval of other data packets through error propagation when using the DNA Fountain algorithm.

Continue reading “Towards practical and robust DNA-based data archiving using the yin–yang codec system” | >

Researchers in Japan have developed a new method for making 5-cm (2-in) wafers of diamond that could be used for quantum memory. The ultra-high purity of the diamond allows it to store a staggering amount of data – the equivalent of one billion Blu-Ray discs.

Diamond is one of the most promising materials for practical quantum computing systems, including memory. A particular defect in the crystal, known as a nitrogen-vacancy center, can be used to store data in the form of superconducting quantum bits (qubits), but too much nitrogen in the diamond disrupts its quantum storage capabilities.

That meant there was a trade-off to make – scientists had to create either large diamond wafers with too much nitrogen, or ultra-pure diamond wafers that are too small to be of much use for data storage. But now, researchers at Saga University and Adamant Namiki Precision Jewelery Co. in Japan have developed a new method for manufacturing ultra-high purity diamond wafers that are big enough for practical use.

Read more | >

A microwave dish transmitter is pointed toward a rectifying antenna in part of the Safe and Continuous Power Beaming – Microwave (SCOPE-M) demonstration at Army Blossom Point Research Field, Maryland, Sept. 21, 2021. U.S. Naval Research Laboratory developed the rectifying antenna, “rectenna”, to convert an x-band microwave beam to 1 kilowatts of DC power at a range of 1 kilometer.

Photo Information

Read more | >

Experimental observations conclude learning is mainly performed by neural … See more.

Summary: Experimental observations conclude learning is mainly performed by neural dendrite trees as opposed to modifying solely through the strength of the synapses, as previously believed.

Source: Bar-Ilan University

The brain is a complex network containing billions of neurons. Each of these neurons communicates simultaneously with thousands of others via their synapses (links), and collects incoming signals through several extremely long, branched “arms,” called dendritic trees.

Continue reading “New Brain Learning Mechanism Calls for Revision of Long-Held Neuroscience Hypothesis” | >

MIT researchers have developed a portable desalination unit, weighing less than 10 kilograms, that can remove particles and salts to generate drinking water.

The suitcase-sized device, which requires less power to operate than a cell phone charger, can also be driven by a small, portable solar panel, which can be purchased online for around $50. It automatically generates drinking that exceeds World Health Organization quality standards. The technology is packaged into a user-friendly device that runs with the push of one button.

Unlike other portable desalination units that require water to pass through filters, this device utilizes to remove particles from drinking water. Eliminating the need for replacement filters greatly reduces the long-term maintenance requirements.

Read more | >

😀


A team of international researchers have identified a genetic cause of lupus. Researchers of the study pinpointed that DNA mutations in a gene that senses viral RNA represents one cause of the chronic condition, affecting approximately 1 in 1,000 people living in the UK. It is important to note that this genetic cause is not the sole trigger for everyone affected by lupus.

Researchers of the study sequenced the whole DNA genome of a juvenile systemic lupus erythematosus (JSLE) patient called Gabriela, who was diagnosed with severe lupus at the age of seven. A severe case such as this, with early onset of symptoms, is a rarity and is commonly associated with a single genetic cause, unlike adult-onset lupus.

The researchers that carried out the genetic analysis identified a single point mutation in the Toll Like Receptor 7 (TLR7) gene. Furthermore the researchers discovered other cases of severe lupus where this gene was also mutated.

Continue reading “Researchers find a genetic cause for lupus” | >

One key aspect of intelligence is the ability to quickly learn how to perform a new task when given a brief instruction. For instance, a child may recognise real animals at the zoo after seeing a few pictures of the animals in a book, despite any differences between the two. But for a typical visual model to learn a new task, it must be trained on tens of thousands of examples specifically labelled for that task. If the goal is to count and identify animals in an image, as in “three zebras”, one would have to collect thousands of images and annotate each image with their quantity and species. This process is inefficient, expensive, and resource-intensive, requiring large amounts of annotated data and the need to train a new model each time it’s confronted with a new task. As part of DeepMind’s mission to solve intelligence, we’ve explored whether an alternative model could make this process easier and more efficient, given only limited task-specific information.

Today, in the preprint of our paper, we introduce Flamingo, a single visual language model (VLM) that sets a new state of the art in few-shot learning on a wide range of open-ended multimodal tasks. This means Flamingo can tackle a number of difficult problems with just a handful of task-specific examples (in a “few shots”), without any additional training required. Flamingo’s simple interface makes this possible, taking as input a prompt consisting of interleaved images, videos, and text and then output associated language.

Similar to the behaviour of large language models (LLMs), which can address a language task by processing examples of the task in their text prompt, Flamingo’s visual and text interface can steer the model towards solving a multimodal task. Given a few example pairs of visual inputs and expected text responses composed in Flamingo’s prompt, the model can be asked a question with a new image or video, and then generate an answer.

Read more | >