Make the test cases conform to VCF specification

Most of these are cosmetic, but some are likely to fix at least some of the platform-dependant test failures reported in #720. Specifically the merging code relies on the correct number of fields in the FORMAT/Number=G tags.
samtools · Dec 13, 2017 · 9a38c20 · 9a38c20
1 parent d945e75
commit 9a38c20
Show file tree

Hide file tree

Showing 51 changed files with 82 additions and 51 deletions.
diff --git a/test/aa.out b/test/aa.out
@@ -1,5 +1,6 @@
 ##fileformat=VCFv4.2
 ##FILTER=<ID=PASS,Description="All filters passed">
+##reference=file://some/path/human_g1k_v37.fasta
 ##contig=<ID=1,length=2147483647>
 ##contig=<ID=2,length=2147483647>
 ##contig=<ID=3,length=2147483647>

diff --git a/test/aa.vcf b/test/aa.vcf
@@ -1,5 +1,6 @@
 ##fileformat=VCFv4.2
 ##FILTER=<ID=PASS,Description="All filters passed">
+##reference=file://some/path/human_g1k_v37.fasta
 ##contig=<ID=1,length=2147483647>
 ##contig=<ID=2,length=2147483647>
 ##contig=<ID=3,length=2147483647>

diff --git a/test/ad-bias.vcf b/test/ad-bias.vcf
@@ -1,5 +1,6 @@
 ##fileformat=VCFv4.2
 ##FILTER=<ID=PASS,Description="All filters passed">
+##reference=file://some/path/human_g1k_v37.fasta
 ##contig=<ID=22,length=51304566>
 ##FORMAT=<ID=AD,Number=R,Type=Integer,Description="Allelic depths">
 ##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">

diff --git a/test/annotate.out b/test/annotate.out
@@ -39,5 +39,5 @@
 2	3199812	indel_3199812	G	GTT,GT	82.7	PASS	AN=4;AC=2,2;INDEL	GT:GQ:DP	1/2:322:26	1/2:322:26
 3	3212016	indel_3212016	CTT	C,CT	79	PASS	AN=4;AC=2,2;INDEL	GT:GQ:DP	1/2:91:26	1/2:91:26
 4	3258448	indel_3258448	TACACACAC	T	59.9	PASS	AN=4;AC=2;INDEL	GT:GQ:DP	0/1:325:31	0/1:325:31
-4	4000000	id1	T	A,C	59.9	PASS	AN=4;AC=2;INDEL	GT:GQ:DP	0/1:325:31	0/1:325:31
+4	4000000	id1	T	A,C	59.9	PASS	AN=4;AC=2,0;INDEL	GT:GQ:DP	0/1:325:31	0/1:325:31
 4	4000001	id2	T	A	59.9	PASS	AN=4;AC=2;INDEL	GT:GQ:DP	0/1:325:31	0/1:325:31
diff --git a/test/annotate.vcf b/test/annotate.vcf
@@ -35,5 +35,5 @@
 2	3199812	.	G	GTT,GT	82.7	PASS	AN=4;AC=2,2	GT:GQ:DP	1/2:322:26	1/2:322:26
 3	3212016	.	CTT	C,CT	79	PASS	AN=4;AC=2,2	GT:GQ:DP	1/2:91:26	1/2:91:26
 4	3258448	.	TACACACAC	T	59.9	PASS	AN=4;AC=2	GT:GQ:DP	0/1:325:31	0/1:325:31
-4	4000000	.	T	A,C	59.9	PASS	AN=4;AC=2	GT:GQ:DP	0/1:325:31	0/1:325:31
+4	4000000	.	T	A,C	59.9	PASS	AN=4;AC=2,0	GT:GQ:DP	0/1:325:31	0/1:325:31
 4	4000001	.	T	A	59.9	PASS	AN=4;AC=2	GT:GQ:DP	0/1:325:31	0/1:325:31
diff --git a/test/annotate2.out b/test/annotate2.out
@@ -39,5 +39,5 @@
 2	3199812	.	G	GTT,GT	82.7	PASS	AN=4;AC=2,2	GT:GQ:DP	1/2:322:26	1/2:322:26
 3	3212016	.	CTT	C,CT	79	PASS	AN=4;AC=2,2	GT:GQ:DP	1/2:91:26	1/2:91:26
 4	3258448	.	TACACACAC	T	59.9	PASS	AN=4;AC=2	GT:GQ:DP	0/1:325:31	0/1:325:31
-4	4000000	.	T	A,C	59.9	PASS	AN=4;AC=2	GT:GQ:DP	0/1:325:31	0/1:325:31
+4	4000000	.	T	A,C	59.9	PASS	AN=4;AC=2,0	GT:GQ:DP	0/1:325:31	0/1:325:31
 4	4000001	.	T	A	59.9	PASS	AN=4;AC=2	GT:GQ:DP	0/1:325:31	0/1:325:31
diff --git a/test/annotate3.out b/test/annotate3.out
@@ -36,5 +36,5 @@
 2	3199812	id11	G	GTT,GT	99	PASS	AN=4;AC=2,2;STR=id11	GT:GQ:DP	1/2:322:26	1/2:322:26
 3	3212016	id12	CTT	C,CT	99	PASS	AN=4;AC=2,2;STR=id12	GT:GQ:DP	1/2:91:26	1/2:91:26
 4	3258448	id13	TACACACAC	T	99	PASS	AN=4;AC=2;STR=id13	GT:GQ:DP	0/1:325:31	0/1:325:31
-4	4000000	.	T	A,C	59.9	PASS	AN=4;AC=2	GT:GQ:DP	0/1:325:31	0/1:325:31
+4	4000000	.	T	A,C	59.9	PASS	AN=4;AC=2,0	GT:GQ:DP	0/1:325:31	0/1:325:31
 4	4000001	.	T	A	59.9	PASS	AN=4;AC=2	GT:GQ:DP	0/1:325:31	0/1:325:31
diff --git a/test/annots.vcf b/test/annots.vcf
@@ -27,7 +27,7 @@
 1	3062915	idSNP	G	T,C	99	PASS	STR=testSNP;TEST=5;DP4=1,2,3,4;AN=3;AC=0,0	GT:TT:GQ:DP:GL	0|0:9,9:999:99:-99,-9,-99,-99,-9,-99	0:9,9:999:99:-99,-9,-99
 1	3106154	id4	CAAA	C	99	PASS	STR=id4;AN=4;AC=0	GT:GQ:DP	0|0:999:99	0|0:999:99
 1	3106154	id5	C	CT	99	PASS	STR=id5;AN=4;AC=0	GT:GQ:DP	0|0:999:99	0|0:999:99
-1	3157410	id6	GA	GC,G	99	PASS	STR=id6;AN=4;AC=0	GT:GQ:DP	0|0:99:99	0|0:99:99
+1	3157410	id6	GA	GC,G	99	PASS	STR=id6;AN=4;AC=0,0	GT:GQ:DP	0|0:99:99	0|0:99:99
 1	3162006	id7	GAA	GG	99	PASS	STR=id7;AN=4;AC=0	GT:GQ:DP	0|0:999:99	0|0:999:99
 1	3177144	id8	G	T	99	PASS	STR=id8;AN=4;AC=0	GT:GQ:DP	0|0:999:99	0|0:999:99
 1	3177144	id9	G	.	99	PASS	STR=id9;AN=4;AC=0	GT:GQ:DP	0|0:999:99	0|0:999:99

diff --git a/test/check.vcf b/test/check.vcf
@@ -12,6 +12,7 @@
 ##contig=<ID=1,assembly=b37,length=249250621>
 ##contig=<ID=3,assembly=b37,length=198022430>
 ##contig=<ID=4,assembly=b37,length=191154276>
+##contig=<ID=2,assembly=b37,length=191154276>
 ##reference=file:///lustre/scratch105/projects/g1k/ref/main_project/human_g1k_v37.fasta
 ##readme=AAAAAA
 ##readme=BBBBBB
@@ -29,8 +30,8 @@
 2	3199812	.	G	GTT,GT	82.7	PASS	AN=4;AC=2,2	GT:GQ:DP	1/2:322:26	1/2:322:26
 3	3212016	.	CTT	C,CT	79	PASS	AN=4;AC=2,2	GT:GQ:DP	1/2:91:26	1/2:91:26
 4	3258448	.	TACACACAC	T	59.9	PASS	DP=62;AN=4;AC=2	GT:GQ:DP	0/1:325:31	0/1:325:31
-4	3258449	.	GCAAA	GA,G	59.9	PASS	DP=62;AN=4;AC=2	GT:GQ:DP	0/1:325:31	0/1:325:31
-4	3258450	.	AAAAGAAAAAG	A,AAAAAAG	59.9	PASS	DP=60;AN=4;AC=2	GT:GQ:DP	0/1:325:31	0/1:325:31
+4	3258449	.	GCAAA	GA,G	59.9	PASS	DP=62;AN=4;AC=2,0	GT:GQ:DP	0/1:325:31	0/1:325:31
+4	3258450	.	AAAAGAAAAAG	A,AAAAAAG	59.9	PASS	DP=60;AN=4;AC=2,0	GT:GQ:DP	0/1:325:31	0/1:325:31
 4	3258451	.	AAA	AGT	59.9	PASS	AN=4;AC=2	GT:GQ:DP	0/1:325:31	0/1:325:31
 4	3258452	.	AAA	AGA	59.9	PASS	AN=4;AC=2	GT:GQ:DP	0/1:325:31	0/1:325:31
 4	3258453	.	AACA	AGA	59.9	PASS	AN=4;AC=2	GT:GQ:DP	0/1:325:31	0/1:325:31

diff --git a/test/consensus.vcf b/test/consensus.vcf
@@ -1,7 +1,10 @@
 ##fileformat=VCFv4.2
 ##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
+##reference=file://some/path/human_g1k_v37.fasta
 ##INFO=<ID=END,Number=1,Type=Integer,Description="End position of the variant described in this record">
 ##ALT=<ID=DEL,Description="Deletion">
+##contig=<ID=1,assembly=b37,length=249250621>
+##contig=<ID=2,assembly=b37,length=249250621>
 #CHROM	POS	ID	REF	ALT	QUAL	FILTER	INFO	FORMAT	NA001
 1	5	.	C	a	.	PASS	.	GT	0/1
 1	5	.	C	t	.	PASS	.	GT	0/1

diff --git a/test/consensus2.vcf b/test/consensus2.vcf
@@ -2,6 +2,11 @@
 ##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
 ##INFO=<ID=END,Number=1,Type=Integer,Description="End position of the variant described in this record">
 ##ALT=<ID=DEL,Description="Deletion">
+##reference=file://some/path/human_g1k_v37.fasta
+##contig=<ID=1,assembly=b37,length=249250621>
+##contig=<ID=2,assembly=b37,length=249250621>
+##contig=<ID=3,assembly=b37,length=249250621>
+##contig=<ID=4,assembly=b37,length=249250621>
 #CHROM	POS	ID	REF	ALT	QUAL	FILTER	INFO	FORMAT	NA001
 1	59	.	C	a,Caa	.	PASS	.	GT	1|2
 2	60	.	C	a,Caa	.	PASS	.	GT	1|2

diff --git a/test/convert.gt.vcf b/test/convert.gt.vcf
@@ -1,6 +1,7 @@
 ##fileformat=VCFv4.1
 ##FILTER=<ID=PASS,Description="All filters passed">
 ##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
+##reference=file://some/path/human_g1k_v37.fasta
 ##contig=<ID=X,assembly=b37,length=155270560>
 ##bcftools_normVersion=1.2+htslib-1.2.1
 ##bcftools_normCommand=norm -m - convert.vcf

diff --git a/test/convert.gvcf.out b/test/convert.gvcf.out
@@ -1,5 +1,6 @@
 ##fileformat=VCFv4.1
 ##FILTER=<ID=PASS,Description="All filters passed">
+##reference=file://some/path/human_g1k_v37.fasta
 ##FORMAT=<ID=GQX,Number=1,Type=Integer,Description="Minimum of {Genotype quality assuming variant position,Genotype quality assuming non-variant position}">
 ##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
 ##FORMAT=<ID=GQ,Number=1,Type=Float,Description="Genotype Quality">

diff --git a/test/convert.gvcf.vcf b/test/convert.gvcf.vcf
@@ -1,5 +1,6 @@
 ##fileformat=VCFv4.1
 ##FILTER=<ID=PASS,Description="All filters passed">
+##reference=file://some/path/human_g1k_v37.fasta
 ##FORMAT=<ID=GQX,Number=1,Type=Integer,Description="Minimum of {Genotype quality assuming variant position,Genotype quality assuming non-variant position}">
 ##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
 ##FORMAT=<ID=GQ,Number=1,Type=Float,Description="Genotype Quality">

diff --git a/test/convert.vcf b/test/convert.vcf
@@ -1,5 +1,6 @@
 ##fileformat=VCFv4.1
 ##FILTER=<ID=PASS,Description="All filters passed">
+##reference=file://some/path/human_g1k_v37.fasta
 ##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
 ##FORMAT=<ID=PL,Number=G,Type=Integer,Description="List of Phred-scaled genotype likelihoods">
 ##FORMAT=<ID=GP,Number=G,Type=Float,Description="Estimated Genotype Probability">

diff --git a/test/csq.vcf b/test/csq.vcf
@@ -1,6 +1,7 @@
 ##fileformat=VCFv4.2
 ##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
-##INFO=<ID=EXP,Number=1,Type=String,Description="Expected consequence">
+##reference=file://some/path/human_g1k_v37.fasta
+##INFO=<ID=EXP,Number=.,Type=String,Description="Expected consequence">
 ##contig=<ID=1,length=2147483647>
 ##contig=<ID=2,length=2147483647>
 ##contig=<ID=3,length=2147483647>

diff --git a/test/empty.vcf b/test/empty.vcf
@@ -3,4 +3,5 @@
 ##contig=<ID=20,length=63025520>
 ##contig=<ID=X,length=155270560>
 ##contig=<ID=Y,length=59373566>
+##reference=file://some/path/human_g1k_v37.fasta
 #CHROM	POS	ID	REF	ALT	QUAL	FILTER	INFO	FORMAT	NA00001	NA00002	NA00003
diff --git a/test/ex2.vcf b/test/ex2.vcf
@@ -7,14 +7,14 @@
 ##INFO=<ID=AA,Number=1,Type=String,Description="Ancestral Allele">
 ##INFO=<ID=DB,Number=0,Type=Flag,Description="dbSNP membership, build 129">
 ##INFO=<ID=H2,Number=0,Type=Flag,Description="HapMap2 membership">
-##INFO=<ID=HOMSEQ,Number=.,Type=String>
+##INFO=<ID=HOMSEQ,Number=.,Type=String,Description="testing">
 ##FILTER=<ID=q10,Description="Quality below 10">
 ##FILTER=<ID=s50,Description="Less than 50% of samples have data">
 ##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
 ##FORMAT=<ID=GQ,Number=1,Type=Integer,Description="Genotype Quality">
 ##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
 ##FORMAT=<ID=HQ,Number=2,Type=Integer,Description="Haplotype Quality">
-##FORMAT=<ID=CNL,Number=.,Type=Integer>
+##FORMAT=<ID=CNL,Number=.,Type=Integer,Description="Some description">
 #CHROM	POS	ID	REF	ALT	QUAL	FILTER	INFO	FORMAT	NA00001	NA00002	NA00003
 20	14370	rs6054257	G	A	29.1	.	NS=3;DP=14;AF=0.5;HOMSEQ;DB	GT:GQ:DP:HQ:CNL	0|0:48:1:25,30:10,20	1|0:48:8:49,51:.	./.:43:5:.,.:1
 20	17330	.	T	A	.	q10;s50	NS=3;DP=11;AF=0.017;H2	GT:GQ:DP:HQ	0|0:49:3:58,50	0|1:3:5:65,3	0/0:41:3:4,5

diff --git a/test/fill-tags.2.out b/test/fill-tags.2.out
@@ -33,7 +33,7 @@
 ##INFO=<ID=AF,Number=A,Type=Float,Description="Allele frequency">
 ##INFO=<ID=MAF,Number=A,Type=Float,Description="Minor Allele frequency">
 #CHROM	POS	ID	REF	ALT	QUAL	FILTER	INFO	FORMAT	NA00001	NA00002	NA00003
-11	2343543	.	A	.	999	PASS	DP=100223;NS=3;AN=6	GT:PL:DP:GQ	0/0:0,255,255:193:99	0/0:0,255,255:211:99	0/0:0,255,255:182:99
+11	2343543	.	A	.	999	PASS	DP=100223;NS=3;AN=6	GT:PL:DP:GQ	0/0:0:193:99	0/0:0:211:99	0/0:0:182:99
 11	5464562	.	C	T	999	PASS	DP=0;NS=0;AN=0;AC=0	GT:PL:DP:GQ	./.:0,0,0:.:.	./.:0,0,0:.:.	./.:0,0,0:.:.
 20	76962	rs6111385	T	C	999	PASS	DP4=110138,70822,421911,262673;DP=911531;Dels=0;FS=21.447;HWE=0.491006;ICF=-0.01062;MQ0=1;MQ=46;PV4=2.5e-09,0,0,1;QD=22.31;NS=3;AN=6;AF=0.833333;MAF=0.166667;AC=5	GT:PL:DP:GQ	0/1:255,0,255:193:99	1/1:255,255,0:211:99	1/1:255,255,0:182:99
 20	126310	.	ACC	A	999	StrandBias;EndDistBias	DP4=125718,95950,113812,80890;DP=461867;HWE=0.24036;ICF=0.01738;INDEL;IS=374,0.937343;MQ=49;PV4=9e-30,1,0,3.8e-13;QD=0.0172;AN=6;AC=4;NS=3;AF=0.666667;MAF=0.333333	GT:DP:GQ:PL	0/1:117:99:255,0,132	0/1:111:99:255,0,139	1/1:78:99:255,213,0

diff --git a/test/fill-tags.3.out b/test/fill-tags.3.out
@@ -32,7 +32,7 @@
 ##INFO=<ID=AC_AB,Number=A,Type=Integer,Description="Allele count in genotypes in AB">
 ##INFO=<ID=AC_BC,Number=A,Type=Integer,Description="Allele count in genotypes in BC">
 #CHROM	POS	ID	REF	ALT	QUAL	FILTER	INFO	FORMAT	NA00001	NA00002	NA00003
-11	2343543	.	A	.	999	PASS	DP=100223	GT:PL:DP:GQ	0/0:0,255,255:193:99	0/0:0,255,255:211:99	0/0:0,255,255:182:99
+11	2343543	.	A	.	999	PASS	DP=100223	GT:PL:DP:GQ	0/0:0:193:99	0/0:0:211:99	0/0:0:182:99
 11	5464562	.	C	T	999	PASS	DP=0;AC_AB=0;AC_BC=0;AC=0	GT:PL:DP:GQ	./.:0,0,0:.:.	./.:0,0,0:.:.	./.:0,0,0:.:.
 20	76962	rs6111385	T	C	999	PASS	DP4=110138,70822,421911,262673;DP=911531;Dels=0;FS=21.447;HWE=0.491006;ICF=-0.01062;MQ0=1;MQ=46;PV4=2.5e-09,0,0,1;QD=22.31;AC_AB=3;AC_BC=4;AC=5	GT:PL:DP:GQ	0/1:255,0,255:193:99	1/1:255,255,0:211:99	1/1:255,255,0:182:99
 20	126310	.	ACC	A	999	StrandBias;EndDistBias	DP4=125718,95950,113812,80890;DP=461867;HWE=0.24036;ICF=0.01738;INDEL;IS=374,0.937343;MQ=49;PV4=9e-30,1,0,3.8e-13;QD=0.0172;AN=6;AC=4;AC_AB=2;AC_BC=3	GT:DP:GQ:PL	0/1:117:99:255,0,132	0/1:111:99:255,0,139	1/1:78:99:255,213,0

diff --git a/test/filter-missing-floats.vcf b/test/filter-missing-floats.vcf
@@ -1,4 +1,5 @@
 ##fileformat=VCFv4.2
+##reference=file:///some/path/1000GenomesPilot-NCBI36.fasta
 ##contig=<ID=1,length=249250621>
 ##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
 ##INFO=<ID=A_AF,Number=A,Type=Float,Description="Allele frequency">

diff --git a/test/filter.10.out b/test/filter.10.out
@@ -1,5 +1,6 @@
 ##fileformat=VCFv4.1
 ##FILTER=<ID=PASS,Description="All filters passed">
+##reference=file:///some/path/1000GenomesPilot-NCBI36.fasta
 ##contig=<ID=chr1,length=135006516>
 ##INFO=<ID=TEST1,Number=1,Type=Integer,Description="Test1">
 ##INFO=<ID=TEST2,Number=1,Type=Float,Description="Test2">

diff --git a/test/filter.4.vcf b/test/filter.4.vcf
@@ -1,4 +1,5 @@
 ##fileformat=VCFv4.1
+##reference=file:///some/path/1000GenomesPilot-NCBI36.fasta
 ##contig=<ID=chr1,length=135006516>
 ##INFO=<ID=TEST1,Number=1,Type=Integer,Description="Test1">
 ##INFO=<ID=TEST2,Number=1,Type=Float,Description="Test2">

diff --git a/test/fixref.2.out b/test/fixref.2.out
@@ -1,5 +1,6 @@
 ##fileformat=VCFv4.2
 ##FILTER=<ID=PASS,Description="All filters passed">
+##reference=file:///some/path/1000GenomesPilot-NCBI36.fasta
 ##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
 ##contig=<ID=1,length=2147483647>
 #CHROM	POS	ID	REF	ALT	QUAL	FILTER	INFO	FORMAT	XY00001

diff --git a/test/fixref.2a.vcf b/test/fixref.2a.vcf
@@ -1,4 +1,5 @@
 ##fileformat=VCFv4.2
+##reference=file:///some/path/1000GenomesPilot-NCBI36.fasta
 ##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
 ##contig=<ID=1,length=2147483647>
 #CHROM	POS	ID	REF	ALT	QUAL	FILTER	INFO	FORMAT	XY00001

diff --git a/test/fixref.2b.vcf b/test/fixref.2b.vcf
@@ -1,4 +1,5 @@
 ##fileformat=VCFv4.2
+##reference=file:///some/path/1000GenomesPilot-NCBI36.fasta
 ##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
 ##contig=<ID=1,length=2147483647>
 #CHROM	POS	ID	REF	ALT	QUAL	FILTER	INFO

diff --git a/test/fixref.vcf b/test/fixref.vcf
@@ -1,4 +1,5 @@
 ##fileformat=VCFv4.2
+##reference=file:///some/path/1000GenomesPilot-NCBI36.fasta
 ##FILTER=<ID=PASS,Description="All filters passed">
 ##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
 ##contig=<ID=1,length=2147483647>

diff --git a/test/gvcf.merge.1.out b/test/gvcf.merge.1.out
@@ -1,10 +1,10 @@
 ##fileformat=VCFv4.1
 ##FILTER=<ID=PASS,Description="All filters passed">
+##reference=file:///some/path/1000GenomesPilot-NCBI36.fasta
 ##FILTER=<ID=LowGQX,Description="Locus GQX is less than 30 or not present">
 ##FILTER=<ID=HighDPFRatio,Description="The fraction of basecalls filtered out at a site is greater than 0.3">
 ##FILTER=<ID=nc,Description="No-call">
-##contig=<ID=chrY,length=59373566,assembly=B37,md5=1e86411d73e6f00a10590f976be01623,species="Homo sapiens">
-##contig=<ID=chrM,length=16569,assembly=B37,md5=c68f52674c9fb33aef52dcf399755519,species="Homo sapiens">
+##contig=<ID=chr1,length=59373566,assembly=B37,md5=1e86411d73e6f00a10590f976be01623,species="Homo sapiens">
 ##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
 ##FORMAT=<ID=HQ,Number=2,Type=Integer,Description="Haplotype Quality">
 ##FORMAT=<ID=GQX,Number=1,Type=Integer,Description="Minimum of {Genotype quality assuming variant position,Genotype quality assuming non-variant position}">
@@ -22,7 +22,6 @@
 ##INFO=<ID=CGA_FI,Number=A,Type=String,Description="Functional impact annotation">
 ##INFO=<ID=END,Number=1,Type=Integer,Description="End position of the region described in this record">
 ##INFO=<ID=BLOCKAVG_min30p3a,Number=0,Type=Flag,Description="Non-variant site block. All sites in a block are constrained to be non-variant, have the same filter value, and have all sample values in range [x,y], y <= max(x+3,(x*1.3)). All printed site block sample values are the minimum observed in the region spanned by the block">
-##contig=<ID=chr1>
 #CHROM	POS	ID	REF	ALT	QUAL	FILTER	INFO	FORMAT	G06	D05	H09
 chr1	10106	.	C	.	0	LowGQX	BLOCKAVG_min30p3a;AN=2	GT:GQX:DP:DPF	./.:.:.:.	0/0:12:5:0	./.:.:.:.
 chr1	10107	.	C	.	0	LowGQX;HighDPFRatio	BLOCKAVG_min30p3a;AN=4	GT:GQX:DP:DPF	.:.:0:1	0/0:12:5:0	0/0:5:2:0

diff --git a/test/gvcf.merge.1.vcf b/test/gvcf.merge.1.vcf
@@ -1,10 +1,10 @@
 ##fileformat=VCFv4.1
+##reference=file:///some/path/1000GenomesPilot-NCBI36.fasta
 ##FILTER=<ID=PASS,Description="All filters passed">
 ##FILTER=<ID=LowGQX,Description="Locus GQX is less than 30 or not present">
 ##FILTER=<ID=HighDPFRatio,Description="The fraction of basecalls filtered out at a site is greater than 0.3">
 ##FILTER=<ID=nc,Description="No-call">
-##contig=<ID=chrY,length=59373566,assembly=B37,md5=1e86411d73e6f00a10590f976be01623,species="Homo sapiens">
-##contig=<ID=chrM,length=16569,assembly=B37,md5=c68f52674c9fb33aef52dcf399755519,species="Homo sapiens">
+##contig=<ID=chr1,length=59373566,assembly=B37,md5=1e86411d73e6f00a10590f976be01623,species="Homo sapiens">
 ##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
 ##FORMAT=<ID=HQ,Number=2,Type=Integer,Description="Haplotype Quality">
 ##FORMAT=<ID=GQX,Number=1,Type=Integer,Description="Minimum of {Genotype quality assuming variant position,Genotype quality assuming non-variant position}">

diff --git a/test/gvcf.merge.2.vcf b/test/gvcf.merge.2.vcf
@@ -1,10 +1,10 @@
 ##fileformat=VCFv4.1
+##reference=file:///some/path/1000GenomesPilot-NCBI36.fasta
 ##FILTER=<ID=PASS,Description="All filters passed">
 ##FILTER=<ID=LowGQX,Description="Locus GQX is less than 30 or not present">
 ##FILTER=<ID=HighDPFRatio,Description="The fraction of basecalls filtered out at a site is greater than 0.3">
 ##FILTER=<ID=nc,Description="No-call">
-##contig=<ID=chrY,length=59373566,assembly=B37,md5=1e86411d73e6f00a10590f976be01623,species="Homo sapiens">
-##contig=<ID=chrM,length=16569,assembly=B37,md5=c68f52674c9fb33aef52dcf399755519,species="Homo sapiens">
+##contig=<ID=chr1,length=59373566,assembly=B37,md5=1e86411d73e6f00a10590f976be01623,species="Homo sapiens">
 ##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
 ##FORMAT=<ID=HQ,Number=2,Type=Integer,Description="Haplotype Quality">
 ##FORMAT=<ID=GQX,Number=1,Type=Integer,Description="Minimum of {Genotype quality assuming variant position,Genotype quality assuming non-variant position}">

diff --git a/test/gvcf.merge.3.vcf b/test/gvcf.merge.3.vcf
@@ -1,10 +1,10 @@
 ##fileformat=VCFv4.1
+##reference=file:///some/path/1000GenomesPilot-NCBI36.fasta
 ##FILTER=<ID=PASS,Description="All filters passed">
 ##FILTER=<ID=LowGQX,Description="Locus GQX is less than 30 or not present">
 ##FILTER=<ID=HighDPFRatio,Description="The fraction of basecalls filtered out at a site is greater than 0.3">
 ##FILTER=<ID=nc,Description="No-call">
-##contig=<ID=chrY,length=59373566,assembly=B37,md5=1e86411d73e6f00a10590f976be01623,species="Homo sapiens">
-##contig=<ID=chrM,length=16569,assembly=B37,md5=c68f52674c9fb33aef52dcf399755519,species="Homo sapiens">
+##contig=<ID=chr1,length=59373566,assembly=B37,md5=1e86411d73e6f00a10590f976be01623,species="Homo sapiens">
 ##FORMAT=<ID=GT,Number=1,Type=String,Description="Genotype">
 ##FORMAT=<ID=HQ,Number=2,Type=Integer,Description="Haplotype Quality">
 ##FORMAT=<ID=GQX,Number=1,Type=Integer,Description="Minimum of {Genotype quality assuming variant position,Genotype quality assuming non-variant position}">

diff --git a/test/idx.vcf b/test/idx.vcf
@@ -1,4 +1,5 @@
 ##fileformat=VCFv4.1
+##reference=file:///some/path/1000GenomesPilot-NCBI36.fasta
 ##contig=<ID=1,length=249250621>
 ##contig=<ID=11,length=135006516>
 ##contig=<ID=20,length=63025520>

diff --git a/test/isec.a.vcf b/test/isec.a.vcf
@@ -6,6 +6,10 @@
 ##FORMAT=<ID=DP,Number=1,Type=Integer,Description="Read Depth">
 ##FORMAT=<ID=GL,Number=G,Type=Float,Description="Genotype Likelihood">
 ##FILTER=<ID=q10,Description="Quality below 10">
+##contig=<ID=1,length=249250621>
+##contig=<ID=2,length=249250621>
+##contig=<ID=3,length=249250621>
+##contig=<ID=4,length=249250621>
 ##test=<xx=A,yy=B,zz=C>
 ##reference=file:///lustre/scratch105/projects/g1k/ref/main_project/human_g1k_v37.fasta
 ##INFO=<ID=AC,Number=A,Type=Integer,Description="Allele count in genotypes">
@@ -16,7 +20,7 @@
 1	3062915	.	GTTT	G	1806	q10	DP=35;DP4=1,2,3,4;AN=2;AC=1	GT:GQ:DP:GL	0/1:409:35:-20,-5,-20
 1	3062915	.	G	T	1806	q10	DP=35;DP4=1,2,3,4;AN=2;AC=1	GT:GQ:DP:GL	0/1:409:35:-20,-5,-20
 1	3106154	.	CAAA	C	1792	PASS	DP=32;AN=2;AC=1	GT:GQ:DP	0/1:245:32
-1	3106154	.	C	T,CT	1792	PASS	DP=32;AN=2;AC=1	GT:GQ:DP	0/1:245:32
+1	3106154	.	C	T,CT	1792	PASS	DP=32;AN=2;AC=1,0	GT:GQ:DP	0/1:245:32
 1	3157410	.	GA	G	628	q10	DP=21;AN=2;AC=2	GT:GQ:DP	1/1:21:21
 1	3162006	.	GAA	G	1016	PASS	DP=22;AN=2;AC=1	GT:GQ:DP	0/1:212:22
 1	3177144	.	GT	G	727	PASS	DP=30;AN=2;AC=1	GT:GQ:DP	0/1:150:30