Do I need to run SplitNCigarReads before ASEReadCounter in GATK 4?

[francescoballesio]

I'm working with RNA-seq data and planning to run ASEReadCounter using GATK 4. In older versions (e.g., GATK 3.8), there was a parameter ALLOW_N_CIGAR_READS that allowed the tool to process spliced reads (with N in the CIGAR string).

Since that option is no longer available in GATK 4, I wanted to confirm if I need to run SplitNCigarReads before running ASEReadCounter in GATK 4 or if ASEReadCounter is able to handle those reads automatically?

If yes, does the split BAM produced by SplitNCigarReads ensure that ASEReadCounter handles junction-spanning reads correctly? I’d also appreciate any tips or best practices related to this step.

Thanks in advance for your help!