Possible issues for large genome (No index is available for this BAM file. and exceeding integer length in GTF file)

[JuiTse]

Dear @mozack ,
Here I am processing the transcriptome for plant species with around 22g genome size. The bam file has already been sliced to per chromosome level rather than all-in-one bam file to match the acceptable bam file size. When processing local alignment by abra2, two issues I encountered.

1. Using the samtools, the bai file cannot be created which arise from exceeding 536,870,912 positions in my bam file.
   As an alternative, I use -c function in samtool to create csi index for my bam file.

However, the following issue regarding "No index is available for this BAM file." still present.
Is there any way for abra2 to execute the csi index format? Or suggestions to solve this issue?

The log file is attached.
abra.log

2. When I try on chromosome 1, the chromosome range of gtf file is too large, with the error message in the log file:
   NumberFormatException: For input string: "2147694110"

Is that possible to enlarge the acceptable "Integer data type" in java?
The log file is attached.
abra.log

I am looking forward to know more, thanks.