Back

A pooled genetic approach provides multimodal access to uncharacterized retinal ganglion cell subtypes

Luong, N. H.; Hougham, O.; Leffler, J.; Sivyer, B.; Wright, K. M.

2026-08-13 neuroscience
10.64898/2026.08.07.743592 bioRxiv
Show abstract

Retinal ganglion cells (RGCs) are the sole output neurons of the retina, responsible for transmitting visual information to the brain. Recent transcriptomic studies have revealed extensive molecular diversity among RGCs that parallels their known morphological and functional heterogeneity. Although large-scale efforts have unified the transcriptomic, morphological, and physiological features for a limited number of RGC subtypes, the majority of molecularly defined types remain poorly characterized. Developing approaches that can be used to identify RGC subtypes in a reliable and reproducible manner is critical for understanding their roles in visual processing. We used a MafbmCherry-2A-Cre mouse line to genetically label four uncharacterized RGC subtypes, along with two well-established -RGC subtypes, and we systematically characterized their molecular markers, central brain targets, dendritic morphologies, and light response properties. Within the population of previously uncharacterized MAFB+ RGCs, we identified two OFF-responsive subtypes which we termed "MAFB-midi-OFF" and "MAFB-asymmetric-OFF", and two ON-OFF-responsive subtypes termed "MAFB-equal bistratified" and "MAFB-unequal bistratified". Notably, no single molecular, morphological, or functional characteristic was sufficient to distinguish all MAFB+ subtypes. Instead, accurate subtype classification emerged only through the integration of multiple complementary features, highlighting the importance of multimodal approaches for defining RGC subtype identity and resolving neuronal diversity within the retina.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.