How do I automate the graphing of these data?

[yangjianhunt]

Hi, everybody,

I have result files generated by blastn which then were sorted based on the second field. A typical file looks like:

360 miR156a
1 miR156a
9 miR156a
1 miR156a
10 miR156a
7 miR156a
1 miR156a
705 miR157a
2 miR157a
1 miR157a
5 miR157a
4 miR157a
67 miR157a
5 miR157a
11 miR157a
2 miR157a
34 miR159
3 miR162
3 miR166a
17 miR166a
4 miR166a
103 miR167a
1 miR167a
... .....

The first column is the deepseq read counts for each unique sequence. The 2nd column is the miR IDs that the sequence was aligns to.
I would like to:
1)
Sum the total read counts for each miR IDs (e.g. for miR156a, sum row1-row7);
Generate a bar graph to show the total read counts for each miR ID.

I have more than 20 files like this. I would like to use an automated way of doing this. The R package came to my minds.
But I have not used R before. Can you guys give me some tips or suggestions as about which R package or tools to use? (I can then learn those and figure out)

2)
If possible, generate a table that summarize all the total reads info from the 20 files.
The table that I would like to have is as follows:

miRID sample1 sample2 sample3 ......... sample 20
miR156 103 300 450 .......... 33
miR157 205 300 ..........
miR167 .....
.... .......


Thanks a lot!!

Jian

[kopi-o]

For 1), the bar plot part is easy in R; just use barplot() !

Summing the counts can be done in a lot of different ways. Here is one that is maybe a bit cryptic but will teach you the table() command. Assume you have the table you pasted in a text file called mirna.txt. Try to run the following in R, with the mirna.txt file in the current working directory:

m <- read.table("mirna.txt")
q <- table(m)
totcounts <- as.numeric(rownames(q)) %*% q
barplot(totcounts)

There are of course more transparent ways of summing the counts, but I'm too lazy to type them out :-)

[yangjianhunt]

Thanks a lot, kopi-o.

This looks awesome. I will try it out.

Jian

[yangjianhunt]

solved

I eventually used:
list.files () function to get all the files
lapply () to achieve processing for multiple functions.
read.table () to read data.frame from each file
tapply (SeqCounts, miRNA, sum) to get a counting for each "class"
write.table () to write data into a file, append=TRUE
also used paste() and cat () to write a name before each appendage.
barplot () to draw polt

It took me a couple of days to learn the introductory basics of R. But it was fun and will be useful in the future I hope.

Again, thanks to Kopi-o for point the way: I haven't learned how to used the table () function yet...But I feel confident to be able to learn it now.