linear table to matrix format - text-processing

Question

I would like to convert a linear table to a matrix format.
My input table looks like this and is called "linear_table.tab":
transcript ortho
Transcript_1 ORTHO_1
Transcript_2 ORTHO_2
Transcript_3 ORTHO_3
Transcript_4 ORTHO_4
Transcript_5 ORTHO_5
Transcript_6 ORTHO_6
Transcript_7 ORTHO_5
Transcript_8 ORTHO_1
Transcript_9 ORTHO_4
Transcript_10 ORTHO_5
Transcript_11 ORTHO_2
Transcript_12 ORTHO_7
Transcript_13 ORTHO_8
Transcript_14 ORTHO_5
Transcript_15 ORTHO_2
Transcript_16 ORTHO_9
what I would like my matrix table to look like:
Transcript_1 Transcript_2 Transcript_3 Transcript_4 Transcript_5 Transcript_6 Transcript_7 Transcript_8 Transcript_9 Transcript_10 Transcript_11 Transcript_12 Transcript_13 Transcript_14 Transcript_15 Transcript_16
Transcript_1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0
Transcript_2 0 0 0 0 0 0 0 0 0 0 1 0 0 0 1 0
Transcript_3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Transcript_4 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0
Transcript_5 0 0 0 0 0 0 1 0 0 1 0 0 0 1 0 0
Transcript_6 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Transcript_7 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0
Transcript_8 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Transcript_9 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0
Transcript_10 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0
Transcript_11 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Transcript_12 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Transcript_13 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Transcript_14 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0
Transcript_15 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Transcript_16 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Here is my code using R:
linear.table <- read.table("linear_table.tab", header=T, sep="\t")
library(reshape2)
dcast(linear.table, transcript~ortho, fill=0)
I get the following output in R:
transcript ORTHO_1 ORTHO_2 ORTHO_3 ORTHO_4 ORTHO_5 ORTHO_6 ORTHO_7 ORTHO_8 ORTHO_9
Transcript_1 ORTHO_1 0 0 0 0 0 0 0 0
Transcript_10 0 0 0 0 ORTHO_5 0 0 0 0
Transcript_11 0 ORTHO_2 0 0 0 0 0 0 0
Transcript_12 0 0 0 0 0 0 ORTHO_7 0 0
Transcript_13 0 0 0 0 0 0 0 ORTHO_8 0
Transcript_14 0 0 0 0 ORTHO_5 0 0 0 0
Transcript_15 0 ORTHO_2 0 0 0 0 0 0 0
Transcript_16 0 0 0 0 0 0 0 0 ORTHO_9
Transcript_2 0 ORTHO_2 0 0 0 0 0 0 0
Transcript_3 0 0 ORTHO_3 0 0 0 0 0 0
Transcript_4 0 0 0 ORTHO_4 0 0 0 0 0
Transcript_5 0 0 0 0 ORTHO_5 0 0 0 0
Transcript_6 0 0 0 0 0 ORTHO_6 0 0 0
Transcript_7 0 0 0 0 ORTHO_5 0 0 0 0
Transcript_8 ORTHO_1 0 0 0 0 0 0 0 0
Transcript_9 0 0 0 ORTHO_4 0 0 0 0 0
I am not sure how to proceed in this aspect using R.

Answer 1

Using awk:
$ cat ortho.awk
NR > 1 {
transcript = $1;
ortho = $2;
i = transcript;
j = ortho;
sub("Transcript_", "", i);
sub("ORTHO_", "", j);
imx[i][j] = 1;
}
END {
for (i in imx) {
for (j in imx) {
omx["Transcript_"+i]["Transcript_"+j] = imx[i][j] == "" ? 0 : 1;
}
}
printf("\t");
for (i in omx) {
printf "\tTranscript%d", i;
}
print "";
for (i in omx) {
printf "Transcript%d", i;
for (j in omx) {
printf "\t%d", omx[i][j];
}
print "";
}
}
Idea is to populate a sparse matrix of 1's, then at the end fill it with 0's in the missing spots. Then print it out.