Demo Web App for Calcium Imaging Time-Lapse Series Analysis
This is a demo app that would crush if you attempted to analyse more than four frames. For larger series analysis, download code at GitHub (click). Running the code as a Shiny app would require e.g. RStudio.
Generation of Correlation Heatmaps Online
If Pearson's correlation is done after detrending, negative values are often generated. The app converts all negatives to NaNs (Not-a-Number).
Detrending is done using the following R language expression and it uses multiple breaking points (bp).
br <- 50 # breaking point
break.points <- seq(from = br, to = dim(dat2)[1], by = br)
dat3 <- dat2 # imported data stored in dat2
for(i in 1:dim(dat2)[2]) { # run each Sample in a dataframe in a loop
dat3[, i] <- detrend(dat2[, i], bp = break.points) # detrend the data
}
dat3[, 1] <- dat2[, 1] ## This line replaces detrended "Time" column with original values
The users have an option to choose heatmaps from the R package pheatmap,
which generates a blue-to-yellow-to-red palette, or from the package gplots
using heatmap.2. The latter uses yellow-to-red as default. Users can also choose
palettes such as bluered or greenred.
Users can also calculate the percentage of correlating pairs out of the total number of all possible pairs in the sample. Finally, one can calculate all datapoints above a certain threshold (default is zero), evaluating e.g. overall sample activity.
Video instructions
How to make a scatter plot overlaid with column graphs online
How to Make a Grouped Graph Combined with Scatter Plot Online
How to Make Stacked Bar Charts Online. How to Generate Percent Stacked Bar Charts on the web