In philliplab/MotifBinner: Bin reads by Motif

Quick summary of final dataset obtained

library(MotifBinner)
library(ggplot2)
library(ape)
consensuses <- report_dat$pb_dat$consensuses

Number of sequences

print(length(consensuses))

Number of ambiguity characters

non_acgt <- NULL
for (i in seq_along(consensuses)){
  x <- sum(consensusMatrix(DNAStringSet(consensuses[[i]]))[-(1:4),])
  non_acgt <- c(non_acgt, x)
}
barplot(table(non_acgt))
print(table(non_acgt))

Outlying sequences

dmat <- fast_stringDist(consensuses, 5)
tot_dist <- apply(as.matrix(dmat), 1, sum)
hist(tot_dist)

PCA of dist matrix

x <- pcoa(dmat)
biplot(x)

Top most outlying sequences

normal_ranks <- rank(tot_dist)
normal_ranks <- normal_ranks[order(normal_ranks)]

rank_comp <- data.frame(seq_name = names(normal_ranks),
           normal_rank = normal_ranks,
           normal_dist = tot_dist[match(names(normal_ranks),
                                        names(tot_dist))])
row.names(rank_comp) <- 1:nrow(rank_comp)
kable(head(rank_comp, 25))
kable(tail(rank_comp, 25))