A small example on how to use mash to make a wrong UPGMA tree

UPGMA_books

# download

wget http://www.gutenberg.org/cache/epub/2265/pg2265.txt -O Hamlet.txt
wget http://www.gutenberg.org/cache/epub/1120/pg1120.txt -o Caesar.txt
wget http://www.gutenberg.org/cache/epub/1120/pg1120.txt -O Caesar.txt
wget http://www.gutenberg.org/cache/epub/1121/pg1121.txt -O as_you_like_it.txt
wget http://www.gutenberg.org/cache/epub/1129/pg1129.txt -O macbeth.txt
wget http://www.gutenberg.org/files/5500/5500-0.txt -O Advancement_of_Learning.txt
wget http://www.gutenberg.org/cache/epub/2434/pg2434.txt -O New_Atlantis.txt
wget http://www.gutenberg.org/files/52190/52190-0.txt -O Ecce_homo.txt
wget http://www.gutenberg.org/files/52263/52263-0.txt -O Twilight_of_Idols.txt
wget http://www.gutenberg.org/files/51356/51356-0.txt -O Birth_of_Tragedy.txt
wget http://www.gutenberg.org/cache/epub/38145/pg38145.txt -O Human_All_Too_Human.txt
wget http://www.gutenberg.org/files/23044/23044-0.txt -O Merry_Wives.txt
wget http://www.gutenberg.org/files/22791/22791-0.txt -O Henry_V.txt

# make all sketches, large enough to not get distance=1
for l in *txt; do ./mash-Linux64-v1.1/mash sketch -a -s 50000 $l; done

# make distances, keep only actual distance (but make sure to look at p-values!)
for l in *msh; do for b in *msh; do ./mash-Linux64-v1.1/mash dist $l $b; done; done | cut -f 1,2,3 > Distance_table.txt

# maybe manually edit titles in Distance_table.txt at this point

# Then run the Rscript
library("phangorn")
a <- read.table("Distance_table.txt", stringsAsFactors=F, sep="\t")
# make into matrix
matrix <- reshape(a, direction="wide", idvar="V2", timevar="V1")
# make into distance matrix, drops labels too for some reason
distance <- as.dist(matrix[,-1], upper=F, diag=F)
# get labels for columns/rows back
attr(distance, "Labels") <- matrix[,1]
# draw tree
png("tree.png")
plot(upgma(distance))
dev.off()
# frolick