hey all !

I currently run splicing analysis from paired-end RNA-seq data with multiple conditions. each condition includes 3 samples , all come from human origin .The aligner I chose is STAR/2.7.10a with ensembel GTF annotation of Grch38 release 105. After filtering the BAM files I run rMATS/4.1.2 comparing each time samples from 2 conditions and filtered the JC files in R to get statisticaly significant events . I am using IGV to reproduce sashimi plots. looking at the splicing events in IGV, the inconsistency between this 2 tools is remarkable , most of the events are not noticable and if so the changes are very small and the number of reads at exons in IGV is much higher then that reported on rMATS tables (even JCEC tables). this leads me to mistrust the rmats results and to be confused about how to continue this process. so my question is how to interpert this inconsitency, how to detect and visualize the strongest events and maybe if there are other tools that do better work than rMATS in which i could use? (already tried DEXseq , dosent seem to be good)

• We started to do RT-PCR validation on some of the skipped exons and it is partially working, with results more similar to IGV than to rMATS output.
• I attach here IGV sashimi plot along with the rMATS results for one event.

I am pretty new to bionformatics and splicing analysis so I would appreciate any help. THANKS :)

rmats results for those 2 samples are: IJC_SAMPLE_1-102 SJC_SAMPLE_1-0 IJC_SAMPLE_2-115 SJC_SAMPLE_26 IncLevel1-1 IncLevel2 -0.782