Bizarrely large END position in the output file

[lgh2624]

Hi.
I ran DECA script below:

deca-submit \
--master local[64] \
--driver-class-path $DECA_JAR \
--conf spark.local.dir=~/tmp \
--conf spark.driver.maxResultSize=0 \
--conf spark.kryo.registrationRequired=true \
--driver-memory 64G \
-- coverage \
-L Twist_Exome_RefSeq_targets_hg38.bed \
-I ../fastq/*/*recalibrated.bam \
-o DECA.RD.txt

deca-submit \
--master local[64] \
--driver-class-path $DECA_JAR \
--conf spark.local.dir=~/tmp \
--conf spark.driver.maxResultSize=0 \
--conf spark.kryo.registrationRequired=true \
--driver-memory 64G \
-- normalize_and_discover \
-min_some_quality 29.5 \
-I DECA.RD.txt \
-o DECA.20200724.gff3

And I got impossible end position from output gff3 file.

chrX    sample0136_S157       DUP     155773872       9223372036854775807     10.582000050240811      .       .       END_TARGET=205696;START_TARGET=205632;Q_SOME=99;Q_START=36;Q_STOP=27;Q_EXACT=11;Q_NON_DIPLOID=99

Why this happen?
I'm waiting for your reply. Thanks.