re-order manuscript

[raynamharris]

Start with an “unbiased” (data-driven) approach. Then move to “biased" (hypothesis-driven) analyses.

1. Compare how different the sub-regions differ in the YOKED groups and what PC’s drive the differences.
2. perhaps compare with the “home cage” data as described previously.
3. Compare how training differs from yoked in each sub region.
4. Compare how “one memory” differs from “two memories” in each subregion.
5. GO analysis to see what differs in the one versus two memory groups (compare with the yoked if appropriate i.e. informative).
6. Were specific memory-maintenance/expression genes changed in trained compared to yoked and one compared to two memory groups, in different brain areas? Prkcz, Prkci, Camk2a, Camk2b, Camk2d, Camk2g, Wwc1, Gria1, Gria2?
7. What gene variations were correlated with variations in memory measures (independent of if there were significantly distinct group differences)?
8. Integrating behavior, physiology and gene expression did the networks of significant correlations differ more between trained and yoked? than between the two yoked and the two trained?
9. Are there “hub” elements that connect memory measures to genes for synaptic structure?

[raynamharris]

Unbiased part 1 and 2. PCA analysis of yoked and home cage animals, looking at CA1, CA3 and DG subfields of the hippocampus. Each circle, square, or triangle represents a unique tissue sample from one mouse.

[raynamharris]

Here are a few figures that compare how training differs from yoked in each sub-region (point 3) and how “one memory” differs from “two memories” in each subregion (point 4).

This table-like figure shows how many genes are differentially expressed between any two treatments. I look at all the data together (top left) and by subregion. The DG specific pattern mirrors the overall pattern. Not much happens in CA3. Most differences are between home cage and everything else.

This heat map of all the samples with highly significant genes in rows shows that DG is almost, but not quite, divided into trained versus yoked and home cage.

If you don't separate by brain region these are the strongest treatment effects. Yes, those are white dots for home cage... i need a new color scheme

[raynamharris]

For point 6, pkmz hasn't shown up in the past, but I do need to include home cage samples and rerun this for other genes, perhaps in a subfield specific manner.

[raynamharris]

Here's yet one more way to look at which groups are most different. a correlation matrix across group average gene expression