Filter vcf file using coordinate of bed , the result is SnpSift had filtered some wrong items, why ?

Question

[duolin1]

when I using snpsift software filter vcf by coordinate of bed, I found the result of vcf have wrong item that non_exist in bed ;

this is bed's coordinate

chr8    29608637    29609316
chr8    29609651    29609846

this is the "wrong" item that I found in result that is filtered by bed

< chr7  123666723   .   AT  A   207.73  PASS    AC=1;AF=0.500;AN=2;BaseQRankSum=-0.749;ClippingRankSum=0.860;DP=28;FS=0.000;MLEAC=1;MLEAF=0.500;MQ=60.39;MQ0=0;MQRankSum=-0.416;QD=7.42;ReadPosRankSum=-0.250;SOR=0.495 GT:AD:DP:GQ:PL  0/1:13,15:28:99:245,0,194

< chr8  29609647    .   TAAAAAGA    T   179.73  PASS    AC=1;AF=0.500;AN=2;BaseQRankSum=0.167;ClippingRankSum=-1.303;DP=23;FS=2.066;MLEAC=1;MLEAF=0.500;MQ=60.00;MQ0=0;MQRankSum=0.702;QD=7.81;ReadPosRankSum=0.634;SOR=1.402   GT:AD:DP:GQ:PL  0/1:16,7:23:99:217,0,976

as you can see ,actually this is "wrong" item because this "chr8...." is not accord with bed's coordinate; by the way, I found the wrong item often happened in the first item of the new "chr"( it can also see in my provided item)

this is my code like this

 cat variants.vcf | java -jar SnpSift.jar intervals my_intervals.bed > variants_intersecting_intervals.vcf

so,my question is why , is it a bug in snpsift???

[Pierre Lindenbaum]

which version of snpsift ?

how about just using bcftools view?

[colindaven]

I'd report it to the SnpSift author.

It is a 7bp deletion which overlaps your specified coordinate so it might make sense depending on how you look at it. I would prefer to have such hits in my results.

[duolin1]

thank you very much. I understand..