In section  5.2 of the diffHic manual (Removing trended biases between libraries), why is the MA plot plotted for two non-replicate samples in the input data?

​input <- c("merged_flox_1.h5", "merged_flox_2.h5", "merged_ko_1.h5", "merged_ko_2.h5")
data <- squareCounts(input, width=1e6, param=param)

<...> 

ab <- aveLogCPM(asDGEList(data))
o <- order(ab)
adj.counts <- cpm(asDGEList(data), log=TRUE)
mval <- adj.counts[,3]-adj.counts[,2]
smoothScatter(ab, mval, xlab="A", ylab="M", main="KO vs. Flox")
fit <- loessFit(x=ab, y=mval)
lines(ab[o], fit$fitted[o], col="red")

When I run this code,  adj.counts is a matrix of the format:

 Sample1   Sample2   Sample3   Sample4

1  1.8188661 2.3363146 1.7343820 1.6612176 
2  2.5522622 3.7050666 2.8758002 2.7878327 
3  4.5796852 5.3595436 4.7326327 4.4682953 

I'm confused: why are we exploring the single trended dispersion for  "merged_ko_1.h5" - "merged_flox_2.h5"?

Why aren't we exploring two plots for "merged_ko_1.h5" - "merged_flox_1.h5" *AND* "merged_ko_2.h5" - "merged_flox_2.h5"???

Firstly, it's not the trended dispersion, it's the trend in the M-value, i.e., log-fold change between libraries. Secondly, if you want to make more plots, feel free to do so. MA plots in RNA-seq/microarray analyses are usually made between all pairs of libraries, but I didn't want to bore people with a whole bunch of plots, so I just showed one plot as a demonstration. I guess I could clarify that in the text.