1) Make a shell script to merge all libraries run on multiple lanes (e.g. merge.sh)
~/bin/gatk-4.0.8.1/gatk MergeSamFiles --INPUT ~/markdup.bam --INPUT ~/markdup.bam --OUTPUT ~/1935_merged_markdup.bam
~/bin/gatk-4.0.8.1/gatk MergeSamFiles --INPUT ~/markdup.bam --INPUT ~/markdup.bam --OUTPUT ~/1936_merged_markdup.bam
...

2) Update readgroup info so the merged files are treated as one (or add it to the previous script)
java -jar /data/lib/java/picard.jar AddOrReplaceReadGroups I=1935_merged_markdup.bam O=1935_merged_markdup_RG.bam RGID=4 RGLB=lib12 RGPL=illumina RGPU=unit1 RGSM=1935_m
java -jar /data/lib/java/picard.jar AddOrReplaceReadGroups I=1936_merged_markdup.bam O=1936_merged_markdup_RG.bam RGID=5 RGLB=lib12 RGPL=illumina RGPU=unit1 RGSM=1936_m
...

3) Write a shell script to index these new bam files (or add it to the previous script)
samtools index 1935_merged_markdup_RG.bam
samtools index 1936_merged_markdup_RG.bam
...

4) Call SNPs
~/bin/gatk-4.0.8.1/gatk HaplotypeCaller -R ~/aws_pseudohap_uniq_apr_kraken2_hardmask_allrepeatmasker.fa -I ~/1935_merged_markdup_RG.bam -I ~/1935_merged_markdup_RG.bam -O HapCaller_RGupdate.vcf