Need suggestion/advice regarding JACUSA pileup

[aayushraman]

I am interested in calling all the mismatches/conversions in my RNA modification dataset that is generated using Illumina TruSeq protocol (dUTP). I am using the JACUSA pileup method for this step. I have questions regarding the JACUSA pileup method:

1. I am interested to know why does the JACUSA pileup requires libraries/samples from two different conditions. Why can't we call on a single library/sample?

2. What if I have more than two conditions? I can run JACUSA mpile for each condition w.r.t to the baseline. Then, what is the best way to concatenate the results across all the conditions and baseline if I used the JACUSA2helper R package so I can visualize the mismatch/substitution rate jointly?

Thanks!

[aayushraman]

any answers or suggestions to this question?

[piechottam]

Sry, for the late response.

Just to be clear. To create your desired output, you are using:

jacusa2.jar pileup ?

1. JACUSA2 supports 1 or 2 conditions - in the next version I will remove the constraint and enable an arbitrary number of conditions.
2. You could treat all your bam files as replicates from 1 condition, e.g.:

Given the following files:

# cond1_repl1.bam, cond1_repl2.bam
# cond2_repl1.bam, cond2_repl2.bam
# Run
jacusa.jar pileup -r cond1_vs_cond2.out cond1_repl1.bam,cond1_repl2.bam cond2_repl1.bam,cond2_repl2.bam

jacusa.jar pileup -r pooled_cond1_cond2.out cond1_repl1.bam,cond1_repl2.bam*,*cond2_repl1.bam,cond2_repl2.bam

Regarding your question how to do it in JACUSA2 helper - I need to test my answer to make sure it works as expected - I'll post the answer here by the upcoming Sunday.