We had a lab meeting on Thursday, and it was my turn to present. Since I’m waiting for data to come through, I didn’t have original research to present, so I did a presentation on mouse vision, which I’ve been reading a lot about recently. I recorded the narration, and if you have Powerpoint, you should be able to hear it and see the original slides. It’s about an hour long. Get it while it’s hot. It covers subjects including:
- The relative advantages and inconveniences of mouse as a model for vision
- Techniques available in mice that are unavailable in primates (calcium imaging, slices, EM, optogenetics, etc.)
- The phylogeny of mice versus primates
- The similarities and differences in the hierarchical organization of the visual system of mice versus primates and carnivores
- The latest news on dorsal and ventral visual streams in mice
- The organization and receptive field properties of mouse V1, with an emphasis on Smith & Häusser (2010)
- The principles of selectivity in inhibitory neurons in V1 from Kerlin et al. (2010) and the relevance of this information for the hexagonal lattice model of V1 selectivity by Ringach
- V1 versus S1 versus M1 circuit diagrams from Kätzel et al. (2010)
I ran a bit out of time, and I mentionned en passant that I thought the statistical evidence in Kätzel et al. was a bit weak. To elaborate further, I thought using an ANOVA was inappropriate to analyze the results presented in Figure 5, because ANOVA tests assume normality while by construction the numbers presented represent sampling from a discrete, positive process, so Poisson statistics or negative binomial seems like a better assumption. Also, the difference between statistically significant may not be itself statistically significant, so saying anything under the threshold is real, everything above it is noise is a little much. It may well be that the differences between S1 and V1 disappear using a more appropriate hierarchical model. I’ll leave this for smarter people to decide (e.g. Andrew Gelman or Liam).
Chris later made some very interesting and relevant comments that unfortunately weren’t captured on audio. He talked about the fact that mouse vision seems particularly useful for answering questions about circuits. I’ve mentioned divisive normalization lately. Chris said that neuroscience really needs to figure out WTF feedback connections do; he worked with cooling loops during his postdoc and got weird results, but more precise instruments could presumably answer this question. Apparently Chris’ old advisor Rick Born is on this trail, at least from an anatomical perspective.
He also wondered why people don’t use squirrels instead of mice, since squirrels are diurnal rodents that see well, and I didn’t have a better answer than “the genetic tools aren’t quite as advanced in squirrel”; perhaps a reader can elaborate.
Wang, Q., Sporns, O., & Burkhalter, A. (2012). Network Analysis of Corticocortical Connections Reveals Ventral and Dorsal Processing Streams in Mouse Visual Cortex Journal of Neuroscience, 32 (13), 4386-4399 DOI: 10.1523/JNEUROSCI.6063-11.2012
Wang, Q., Gao, E., & Burkhalter, A. (2011). Gateways of Ventral and Dorsal Streams in Mouse Visual Cortex Journal of Neuroscience, 31 (5), 1905-1918 DOI: 10.1523/JNEUROSCI.3488-10.2011
Smith SL, & Häusser M (2010). Parallel processing of visual space by neighboring neurons in mouse visual cortex. Nature neuroscience, 13 (9), 1144-9 PMID: 20711183
Kerlin, A., Andermann, M., Berezovskii, V., & Reid, R. (2010). Broadly Tuned Response Properties of Diverse Inhibitory Neuron Subtypes in Mouse Visual Cortex Neuron, 67 (5), 858-871 DOI: 10.1016/j.neuron.2010.08.002
Kätzel, D., Zemelman, B., Buetfering, C., Wölfel, M., & Miesenböck, G. (2010). The columnar and laminar organization of inhibitory connections to neocortical excitatory cells Nature Neuroscience, 14 (1), 100-107 DOI: 10.1038/nn.2687
One response to “A presentation on mouse vision”
Squirrels are expensive, take a lot of space, have no genetic tools and are more difficult to use in 2-P microscopy, just to name a few reasons. Hence, there are very little basic data on them (where to record from, what to expect in which layer, etc.), so people don’t want to start from scratch. This also means that vivariums aren’t set up to have them, adding to the activation energy needed to start a squirrel project.
Most scientists want to do what other scientists have done before, with epsilon change. Haven’t you heard of Elmo’s song?