The Science & Technology desk gathers a weekly digest with impactful and interesting research publications and developments at Stanford. Read the latest in this week’s Research Roundup.
Replicating a pain transmission pathway
As detailed in a Stanford-led study published in Nature, researchers successfully created a replica of a neuronal pathway responsible for pain transmission in a lab dish called an “assembloid.”
The pathway, known as the ascending sensory pathway, involves a web of neurons sending sensory information all the way from the ganglions — typically found on our hands — to the cortex. With a replica of this pathway, the researchers were able to both observe the neuronal chemical activity and modulate it by affecting gene expression and other chemical methods.
Why is this important? Chronic pain affects millions of people, and many treatments — especially opioids — are addictive. With a better understanding of how the ascending sensory pathway functions and how it can be modulated, researchers could develop more effective treatments to ameliorate chronic pain.
Psychiatry and behavioral sciences professor Sergiu Pasca, senior author of the study, offered his perspective on the future applications of the research.
“We think screening for drugs that tame sensory organoids’ ability to trigger excessive or inappropriate waves of neuronal transmission through our assembloid, without affecting the brain’s reward circuitry as opioid drugs do — which is why they’re addictive — could lead to better-targeted therapies for pain,” Pasca told Stanford Medicine.
AI and Digital Twins
ChatGPT is an example of a foundational model, in which AI is able to analyze a large amount of text and apply its analysis to other contexts. In a Stanford-affiliated study published in Nature, researchers used an AI algorithm similar to a foundational model to analyze videos of mice watching multiple movies. The model successfully predicted the mice’s neuronal reactions to other visual imagery, acting as a “digital twin.”
Since mice, like humans, have low-resolution vision, their visual system tends to focus more on action compared to finer details such as imagery. To gain a better understanding of mice’s neuronal system, the researchers deliberately chose action movies with extensive movement to properly stimulate them — otherwise, the algorithm would be relying on low-yield data and, more crucially, data that does not represent how mice readily interpret visual information.
Given the scope of the neuronal activity in the brain, it would seem almost impossible to make such predictions. The researchers attribute this predictive capacity to having an extremely large data set. This approach will be especially important for gaining deeper insight into the brain on a neuronal level, according to Andreas Tolias, a senior author of the study and professor of ophthalmology.
“We’re trying to open the black box, so to speak, to understand the brain at the level of individual neurons or populations of neurons and how they work together to encode information,” Tolias told Stanford Medicine.