Tyler O'Neal, Staff Editor CLOUD

Machine learning tool estimates extinction risk for reptiles previously unprioritized for conservation

The iconic Red List of Threatened Species, published by the International Union for Conservation of Nature (IUCN), identifies species at risk of extinction. A study in PLOS Biology publishing May 26th by Gabriel Henrique de Oliveira Caetano at Ben-Gurion University of the Negev, Israel, and colleagues presents a novel machine learning tool for assessing extinction risk and then use this tool to show that reptile species which are unlisted due to lack of assessment or data are more likely to be threatened than assessed species. Potamites montanicola, classified as ‘Critically Endangered’ by automated the assessment method and as ‘Data Deficient’ by the IUCN Red List of Threatened Species. CREDIT Germán Chávez, Wikimedia Commons (CC-BY 3.0, https://creativecommons.org/licenses/by/3.0)

The IUCN’s Red List of Threatened Species is the most comprehensive assessment of the extinction risk of species and informs conservation policy and practices globally. However, the process for categorizing species is laborious and subject to bias, depending heavily on manual curation by human experts; many animal species have therefore not been evaluated, or lack sufficient data, creating gaps in protective measures.

To assess 4,369 reptile species that were previously unable to be prioritized for conservation and develop accurate methods for assessing the extinction risk of obscure species, these researchers created a machine learning supercomputer model. The model assigned IUCN extinction risk categories to the 40% of the world’s reptiles that lacked published assessments or are classified as “DD” (“Data Deficient”) at the time of the study. The researchers validated the model’s accuracy, comparing it to the Red List risk categorizations.

The researchers found that the number of threatened species is much higher than reflected in the IUCN Red List and that both unassessed (“Not Evaluated” or “NE”) and Data Deficient reptiles were more likely to be threatened than assessed species. Future studies are needed to better understand the specific factors underlying extinction risk in threatened reptile taxa, to obtain better data on obscure reptile taxa, and to create conservation plans that include newly identified threatened species.

According to the authors, “Altogether, our models predict that the state of reptile conservation is far worse than currently estimated and that immediate action is necessary to avoid the disappearance of reptile biodiversity. Regions and taxa we identified as likely to be more threatened should be given increased attention in new assessments and conservation planning. Lastly, the method we present here can be easily implemented to help bridge the assessment gap on other less known taxa”.

Coauthor Shai Meiri adds, “Importantly, the additional reptile species identified as threatened by our models are not distributed randomly across the globe or the reptilian evolutionary tree. Our added information highlights that there are more reptile species in peril – especially in Australia, Madagascar, and the Amazon basin – all of which have a high diversity of reptiles and should be targeted for the extra conservation efforts. Moreover, species-rich groups, such as geckos and elapids (cobras, mambas, coral snakes, and others), are probably more threatened than the Global Reptile Assessment currently highlights, these groups should also be the focus of more conservation attention”

Coauthor Uri Roll adds, “Our work could be very important in helping the global efforts to prioritize the conservation of species at risk – for example using the IUCN red-list mechanism. Our world is facing a biodiversity crisis, and severe man-made changes to ecosystems and species, yet funds allocated for conservation are very limited. Consequently, it is key that we use these limited funds where they could provide the most benefits. Advanced tools- such as those we have employed here, together with accumulating data, could greatly cut the time and cost needed to assess extinction risk, and thus pave the way for more informed conservation decision making.”

Rockefeller bioinformatic prospecting, synthesis of antibiotics helps turn the tide against drug-resistant pathogens

Tyler O'Neal, Staff Editor CLOUD

A new antibiotic, synthesized at The Rockefeller University in New York City and derived from supercomputer models of bacterial gene products, appears to neutralize even drug-resistant bacteria. The compound, named cilagicin, works well in mice and employs a novel mechanism to attack MRSA, C. diff, and several other deadly pathogens, according to a study published in Science. The synthetic antibiotic cilagicin was particularly active against Gram-positive bacteria such as Streptococcus pyogenes, depicted above.

The results suggest that a new generation of antibiotics could be derived from computational models. "This isn't just a cool new molecule, it's a validation of a novel approach to drug discovery," says Rockefeller's Sean F. Brady. "This study is an example of computational biology, genetic sequencing, and synthetic chemistry coming together to unlock the secrets of bacterial evolution."

Acting on eons of bacterial warfare

Bacteria have spent billions of years evolving unique ways to kill one another, so it's perhaps unsurprising that many of our most powerful antibiotics are derived from bacteria themselves. With the exceptions of penicillin and a few other notables derived from fungi, most antibiotics were first weaponized by bacteria to fight off fellow bacteria.

"Eons of evolution have given bacteria unique ways of engaging in warfare and killing other bacteria without their foes developing resistance," says Brady, the Evnin Professor and head of the Laboratory of Genetically Encoded Small Molecules. Antibiotic drug discovery once largely consisted of scientists growing streptomyces or bacillus in the lab and bottling their secrets to treat human diseases.

But with the rise of antibiotic-resistant bacteria, there is an urgent need for new active compounds—and we may be running out of bacteria that are easy to exploit. Untold numbers of antibiotics, however, are likely hidden within the genomes of stubborn bacteria that are tricky or impossible to study in the lab. "Many antibiotics come from bacteria, but most bacteria can't be grown in the lab," Brady says. "It follows that we're probably missing out on most antibiotics."

An alternative method, championed by the Brady lab for the past fifteen years, involves finding antibacterial genes in soil and growing them within more lab-friendly bacteria. But even this strategy has its limitations. Most antibiotics are derived from genetic sequences locked within clusters of bacterial genes, known as biosynthetic gene clusters, that function as a unit to collectively code for a series of proteins. But those clusters are often inaccessible with current technologies.

"Bacteria are complicated, and just because we can sequence a gene doesn't mean we know how the bacteria would turn it on to produce proteins," Brady says. "There are thousands and thousands of uncharacterized gene clusters, and we have only ever figured out how to activate a fraction of them."

A new pool of antibiotics

Frustrated with their inability to unlock many bacterial gene clusters, Brady and colleagues turned to algorithms. By teasing apart the genetic instructions within a DNA sequence, modern algorithms can predict the structure of the antibiotic-like compounds that a bacterium with these sequences would produce. Organic chemists can then take that data and synthesize the predicted structure in the lab.

It may not always be a perfect prediction. "The molecule that we end up with is presumably, but not necessarily, what those genes would produce in nature," Brady says. "We aren’t concerned if it is not exactly right—we only need the synthetic molecule to be close enough that it acts similarly to the compound that evolved in nature."

Postdoctoral associates Zonggiang Wang and Bimal Koirala from the Brady lab began by searching through an enormous genetic-sequence database for promising bacterial genes that were predicted to be involved in killing other bacteria and hadn't been examined previously. The "cil" gene cluster, which had not yet been explored in this context, stood out for its proximity to other genes involved in making antibiotics. The researchers duly fed its relevant sequences into an algorithm, which proposed a handful of compounds that cil likely produces. One compound, aptly dubbed cilagicin, turned out to be an active antibiotic.

Cilagicin reliably killed Gram-positive bacteria in the lab, did not harm human cells, and (once chemically optimized for use in animals) successfully treated bacterial infections in mice. Of particular interest, cilagicin was potent against several drug-resistant bacteria and, even when pitted against bacteria grown specifically to resist cilagicin, the synthetic compound prevailed.

Brady, Wang, Koirala, and colleagues determined that cilagicin works by binding two molecules, C55-P and C55-PP, both of which help maintain bacterial cell walls. Existing antibiotics such as bacitracin bind one of those two molecules but never both, and bacteria can often resist such drugs by cobbling together a cell wall with the remaining molecule. The team suspects that cilagicin's ability to take both molecules offline may present an insurmountable barrier that prevents resistance.

Cilagicin is still far from human trials. In follow-up studies, the Brady lab will perform further syntheses to optimize the compound and test it in animal models against more diverse pathogens to determine which diseases it may be most effective in treating.

Beyond the clinical implications of cilagicin, however, the study demonstrates a scalable method that researchers could use to discover and develop new antibiotics. "This work is a prime example of what could be found hidden within a gene cluster," Brady says. "We think that we can now unlock large numbers of novel natural compounds with this strategy, which we hope will provide an exciting new pool of drug candidates.”

Foresight Williams Technology Funds invests in dRISK

Tyler O'Neal, Staff Editor CLOUD
  • Foresight Williams Technology leads the £1.7m super-seed round
     
  • Funding will support further expansion of the business, an autonomous vehicle testing and training company
     
  • dRISK’s core technology uses networks of data to store, visualize and reveal “unknown unknowns” in complex and highly sparse data areas

Foresight Williams Technology (“FWT”) Funds has announced a £1 million investment in dRISK, an autonomous vehicle testing and training company. 1189806805 huge min 1 a4653

Founded in 2014, by CEO Chess Stetson, dRISK is an AI company that promises to revolutionize autonomous vehicles’ (“AVs”) safety by training them to avoid high-risk scenarios. This contrasts with the conventional approach of training AVs on low-risk driving. dRISK counts multiple leading AV developers among its current customer base - which is growing quickly.

dRISK‘s core technology - with four patents granted and two pending - uses networks of data to store, visualize, and reveal “unknown unknowns” in complex and highly sparse data areas.

The commercial rollout of AVs has failed to live up to expectations due to the technical challenge of dealing with high-risk scenarios which are unlikely but pose a real risk. dRISK solves this problem by fusing public, private, and proprietary data covering high-risk scenarios. dRISK’s statistical robustness and auditability have allowed it to win the largest grant from the UK’s Centre for Connected and Autonomous vehicles; enabling dRISK to build the ultimate test for self-driving cars.

The global market for AI training data alone is currently valued at $1.5 billion. FWT’s investment will accelerate the development and growth of dRISK as a central technology for AV development.

Matthew Burke, Head of Technology Ventures at WAE, commented: “dRISK’s proprietary technology has the potential to accelerate the development of automated driving control systems and overcome one of the main barriers to self-driving: the identification of edge cases. We are delighted that FWT has made this investment and we expect to leverage our knowledge of the automotive industry to increase uptake of dRISK’s technology.”

Chess Stetson, Chief Executive of dRISK, added: “This investment will allow dRISK to accelerate our business plan and achieve our goal of introducing the dRISK product to the majority of the AV industry. And in turn, dramatically accelerating the development of safe and useable autonomous vehicles. We thank Foresight and Williams Advanced Engineering for their support.”

Commenting on the investment, Hugh Minnock, Senior Investment Manager at Foresight, said: “Foresight Williams Technology is delighted to invest in dRISK, a hugely exciting company that will hasten the Autonomous Vehicle revolution and the decarbonized future we expect to follow.”