Hi all,

I have different paired-end exome sequencing samples and I would like to call CNVs for them using CoNIFER (http://conifer.sourceforge.net/). Previously, I have to map my samples, and the developer recommends mrsFAST.

1. First of all, I have read (https://sourceforge.net/p/mrfast/discussion/946649/thread/98ef8f35/) that, for calling CNVs, you do not want to use the paired-end mode of the program, but I still do not see clearly why should not I. Any ideas?

2. On the other hand, I have noticed that mrsFAST cannot handle reads with irregular lenght. Am I right? I had to use the --crop command to cut them and make them the same lenght.

3. I am not sure if I am doing everything OK. After creating my bam files, I use CoNIFER to call variants (by creating the RPKM files and the .hdf5 file needed) but I do not see what I expected. In fact, there is a duplication that we have previously seen by experimental techniques that does not appear, so I am afraid that I am missing something and that I messed it up somewhere.

This is my command:

mrsfast --search $hg19 --seq $input/S${i}_R1.fastq${a} --seq $input/S${i}_R2.fastq${a} --mem 25 --crop 130 -o $bams/S${i}_${a}.sam

Any ideas and comments are welcome. :)

Thank you in advance.