Clowney Lab doctoral candidate Maria Ahmed is one of twelve graduate students from around the country honored by being named a 2023 Harold Weintraub Award recipient. She will be presenting her research on how the sense of smell develops at the Weintraub Symposium in May in Seattle.
You can probably distinguish the favorite teas of your colleagues while still recognizing that any tea pales in comparison to coffee, she says. "My work is focused on understanding how the brain is able to perceive the almost limitless sensory information around us, with only a limited set of neurons."
"One of the reasons we can perceive broadly is because of neural circuits called 'expansion layers,' where large groups of neurons receive limited and random combinations of sensory inputs. Expansion layers are found in three of the major clades of animals: the mushroom body of arthropods, the parallel lobe system of cephalopods, and the cerebellum and pallium in vertebrates."
"The quantitative specifics of expansion layers are thought to structure perception. These include how many neurons are in the expansion layer, how many sensory “channels” (pieces of information) are coming in, and how many of these inputs individual neurons receive. While these structures have been the focus of a rich body of theoretical work, since Marr-Albus theory in the 1970s, it has not been possible to test these models."
"In my PhD work, I have developed a suite of methods to change the circuit features of the expansion layer neurons in vivo in the Drosophila melanogaster perception and learning brain center called the mushroom body, and explicitly test the effect of these manipulations on circuit function. Specifically, we changed two wiring variables--the number of expansion layer neurons, and the number of inputs to these neurons--and then tested the olfactory perception and associative learning abilities of these altered-circuit flies."
"Through this work, I aimed to confirm the Marr-Albus theory that the number of inputs to expansion layer neurons determines the diversity of sensory representations. We observed remarkable developmental robustness in the altered circuits, highlighting that the genetic algorithms that set up these circuits prioritise cellular relationships rather than absolute numbers. Ultimately, we hope this work provides a novel method for testing the "purpose" of observed learning architectures."
She will be presenting work in this article: Hacking brain development to test models of sensory coding
The award is supported by the Fred Hutchinson Cancr Center's Weintraub/Groudine Fellowship for Science and Human Disease, which was established to foster intellectual exchange through programs for graduate students, fellows and visiting scholars.