Experimental design edits

A couple more minor tweaks to work on while I’m thinking about them. First, I wanted to modify the experimental temperature figure. I returned to this R Markdown code to update the x-axis. Specifically, I wanted to add all of the dates onto the x-axis. After poking around, I learned that I needed to use scale_x_datetime to add the date information. I could then specify different “breaks” for how frequent the x-axis should be labelled. To get the full range of dates, I used 24 hour breaks:

geom_ribbon(aes(x = dateTime, y = avgColdTemp, ymin = coldMean - coldSD, ymax = coldMean + coldSD), fill = plotColors[3], alpha = 0.15) +
  geom_line(aes(x = dateTime, y = avgColdTemp), color = plotColors[3]) +
  geom_hline(yintercept = 5.7, colour = plotColors[3], linetype = 3) +
  labs(x = "", y = "") +
  scale_x_datetime(date_breaks = "1 day", date_labels = "%B %d") +
  scale_y_continuous(limits = c(5, 7.5), breaks = seq(5, 7.5, 0.5)) +
  theme_classic(base_size = 15) + theme(axis.text.x = element_text(angle = 40, vjust = 1.0, hjust = 1.0),
                                        plot.margin = margin(2,0,1,0))

Initially, I thought I would change the x-axis to only include experimental day, but Carolyn suggested keeping the original months/days because it highlights that the experiment was conducted in the summer. She suggested adding a secondary x-axis for this figure with experimental day. Then I began another deeper rabbit hole dive to figure out how to add a secondary x-axis to a plot! While I could use sec.axis within scale_x_datetime to specify a second axis, this second axis had to be some kind of derivative of the first. And experimental day couldn’t really be derived from date (or, at least I couldn’t figure out a way to do this). So I settled for some post-hoc InDesign modification (I am paying for the Adobe Creative Suite so I might as well use it).

Figure 1. Updated experimental temperature figure

While I had InDesign open, I slapped together an experimental timeline. I wanted to highlight the different timepoints where I measured TTR, did respirometry, or took heart samples for molecular analyses.

Figure 2. Experimental timeline

Finally, I added some details about ESL filtering and brand names for my temperature controllers and immersion heaters.

Guess I gotta go back into metabolomic and lipidomics land and figure more of that stuff out!

Going forward

1. Repeat lipidomics WCNA with temperature and day separately
2. Look into MetaboAnalyst discussion comments
3. Determine how to get more common lipid names that match MetaboAnalyst formatting
4. Reach out to Shelly to see if she tried anything else for lipid analysis
5. Try RaMP enrichment instead of KEGG for lipid sets
6. Normalize for “captivity” effect
7. Look into individual lipid classes from earlier papers
8. Determine if I can add physiology data to mixOmics analyses
9. Integrate metabolomics and lipidomics data