IGV browser display RNA-seq

Question

[Luca]

Hi everyone. I'm analyzing some RNA-seq data. So i as usual generated the index with rsem :

$RSEM/rsem-prepare-reference -p $THREADS --star --star-path /bin/STAR_2.5 --gtf $SCRATCH/genome.gtf $SCRATCH/genome.fa $SCRATCH/genome

then i aligned my fastq to this genome with

$STAR/STAR --genomeDir $REFGEN --readFilesIn $FQ1 $FQ2 --runThreadN $THREADS

then i sort and transform to bam.

java -Djava.io.tmpdir=$DATA/tmp -jar $PICARD/picard.jar SortSam INPUT=Aligned.out.sam OUTPUT=sorted_reads.bam SORT_ORDER=coordinate

java -Djava.io.tmpdir=$DATA/tmp -jar $PICARD/picard.jar BuildBamIndex INPUT=sorted_reads.bam

Now if i transform this bam in coverage to be displayed in igv browser i have such a picture

but if i load directly the bam file on igv i have the following picture (The grey track)

I dont understand why but the BW file act in some weird way on introns and i dont think it is a good way to just remove the introns from the analysis. Do you have any idea of why is is happening? Probably the problem is in how i generate the bedgraph file but i use the standard command from bamtools

bedtools genomecov -ibam bamfile.bam -bg > file.bedgraph

THank you

[jv]

I'm not sure, but I think you need to use the -split flag with genomecov which is provided for working with spliced alignments https://bedtools.readthedocs.io/en/latest/content/tools/genomecov.html#split-reporting-coverage-with-spliced-alignments-or-blocked-bed-features

You might also want to review the following post for some additional insights into making coverage tracks from bam files: https://www.biostars.org/p/92935/#92935