SelectVariants for deletions that start outside interval, but actually overlap the interval
Say if I made calls for an exome and used a padded bed file and got a deletion call like below;
2 241808825 rs140694338 AGGCCTCCCT A 26496.73 PASS AC=1;AF=0.500;AN=2;BaseQRankSum=-0.607;DB;DP=293;FS=0.000;GC=64.84;HaplotypeScore=44.9377;MLEAC=1;MLEAF=0.500;MQ=33.24;MQ0=0;MQRankSum=-2.931;QD=90.43;RPA=2,1;RU=GGCCTCCCT;ReadPosRankSum=-0.308;STR GT:AD:DP:GQ:PL 0/1:3,245:293:99:26534,0,1061
but I end up using select variants with -L 2:241808827-241808900 for whatever reason (say 241808827 is the actual start of my exon). In reality, the deletion above does overlap with the new interval, but SelectVariants would not extract this variant while if I were to use tabix like;
tabix -h foo.vcf.gz 2:241808827-241808900
It would actually extract the deletion call.
I'm guessing that this isn't a bug in GATK, more of a feature, and didn't see anything on here that talks about this, but would just like to hear your thoughts on this