Biology is an area of science where AI and machine-learning approaches have seen some spectacular successes, such as designing enzymes to digest plastics and proteins to block snake venom. But in an era of seemingly endless AI hype, it might be easy to think that we could just set AI loose on the mounds of data we've already generated and end up with a good understanding of most areas of biology, allowing us to skip a lot of messy experiments and the unpleasantness of research on animals.
But biology involves a whole lot more than just protein structures. And it's extremely premature to suggest that AI can be equally effective at handling all aspects of biology. So we were intrigued to see a study comparing a set of AI software packages designed to predict how active genes will be in cells exposed to different conditions. As it turns out, the AI systems couldn't manage to do any better than a deliberately simplified method of predicting.
The results serve as a useful caution that biology is incredibly complex, and developing AI systems that work for one aspect of it is not an indication that they can work for biology generally.
© BSIP
For many species, producing an embryo is a bit of a contest between males and females. Males want as many offspring as possible and want the females to devote as many resources as possible to each of them. Females do better by keeping their options open and distributing resources in a way to maximize the number of offspring they can produce over the course of their lives.
In mammals, this plays out through the chemical modification of DNA, a process called imprinting. Males imprint their DNA by adding methyl modifications to it in a way that alters the activity of genes in order to promote the growth of embryos. Females do similar things chemically but focus on shutting down genes that promote embryonic growth. In a handful of key regions of the genome, having only the modifications specific to one sex is lethal, as the embryo can't grow to match its stage of development.
One consequence of this is that you normally can't produce embryos using only the DNA from eggs or from sperm. But over the last few years, researchers have gradually worked around the need for imprinted sites to have one copy from each parent. Now, in a very sophisticated demonstration, researchers have used targeted editing of methylation to produce mice from the DNA of two sperm.
© David A. Northcott
Over the last century or more, humanity has been developing an ever-growing number of new chemicals that have never been seen before by Earth's creatures. Many of these chemicals end up being toxic contaminants that we'd love to get rid of, but we struggle to purify them from the environment or break them down once we do. And microbes haven't had much chance to evolve the ability to break them down for us.
Over the last few years, however, we've found a growing number of cases where bacteria have evolved the ability to break down such chemicals, like industrial contaminants and plastics. Unfortunately, these bacteria are all different species, target different individual contaminants, and thrive in different environments. But now, researchers have developed a new way to take the genes from all these species and place them in a single bacterial strain that can decontaminate complex waste mixtures.
The inspiration for this work was the fact that a lot of this industrial contamination contains a mixture of toxic organic molecules that are commonly found in brackish or salty water. So, the research team, based in Shenzhen, China, started by simply testing a number of lab bacteria strains to develop one that could survive these conditions. The one that seemed to survive the best was Vibrio natriegens. These bacteria were discovered in a salt marsh, and their primary claim to fame is an impressive growth rate, with a population being able to double about every 10 minutes.
© Aldo Pavan
Even the tiniest of living things are capable of some amazing forms of locomotion, and some come with highly sophisticated sensor suites and manage to source their energy from the environment. Attempts to approach this sort of flexibility with robotics have taken two forms. One involves making tiny robots modeled on animal behavior. The other involves converting a living creature into a robot. So far, either approach has involved giving up a lot. You're either only implementing a few of life's features in the robot or shutting off most of life's features when taking over an insect.
But a team of researchers at Harvard has recognized that there are some behaviors that are so instinctual that it's possible to induce animals to act as if they were robotic. Or mostly robotic, at least—the fruit flies the researchers used would occasionally go their own way, despite strong inducements to stay with the program.
The first bit of behavior involved Drosophila's response to moving visual stimuli. If placed in an area where the fly would see a visual pattern that rotates from left to right, the fly will turn to the right in an attempt to keep the pattern stable. This allowed a projector system to "steer" the flies as they walked across an enclosure (despite their names, fruit flies tend to spend a lot of their time walking). By rotating the pattern back and forth, the researchers could steer the flies between two locations in the enclosure with about 94 percent accuracy.
© arlindo71
Login