lh3 / bgt Goto Github PK

HI!
when i try to run compress order:
./bgt import -S -o archive input.vcf
Error occured

It seems that BGT does not work with joint called variants by DeepVariant/GLNexus process, when run import, it gives this error:

[W::vcf_parse1] FORMAT 'RNC' is not defined in the header

I understand that this is probably caused by the new "RNC" field by GLNexus, but is there any possible walk-around solutions?

Do you have test scripts for bgt?

Could you provide efficient query across the annotations using a FM-index over the concatenated annotation strings from the VCF file? A second compressed bitvector could encode variant annotation starts in this record (basically storing a variant to annotation mapping).

Then you could subset to a given set of records with a particular annotation by finding the ranks of the occurrences of a given pattern in the auxiliary bitvector.

I guess this wouldn't help much when you have to compare floats in the annotations and the annotation is included in all records. Then you end up needing to compare lots of values to execute the query. There might also be a way around this though.

Is it possible to merge bgt files after bgt indexing or should that be done before?

In the BGT example, on my server, 0.0.0.0 does not work:

curl -s '0.0.0.0:8000' | less -S  # help
curl -s '0.0.0.0:8000/?a=(impact=="HIGH")&s=(population=="FIN")&f=(AC>0)'

I switched it to 127.0.0.1 and then the example worked.

I am trying to find insertions using the -a,#variant# framework but I can't get it to return anything. Deletions and SNPs work fine with this format, but insertions always return null. I've tried many variations of the formatting. I even tried to find insertions using the variant IDs generated by the vep2sql.js (using the -f flag) script, but even these return null. Searching for these same variants using the -r region flag works perfectly, so I know they exist in the database. Any help? I don't think its an error with my formatting and it seems limited to insertions.

For example:

This is a variant identified by VEP running the 1kg11-1M.raw.bcf with the following command:
"variant_effect_predictor.pl -i 1kg11-1M.raw.vcf -o 1kg11-1M.raw.vcf.vep --refseq --dir /home/ec2-user/.vep --offline --pick --cache --everything --force_overwrite"
The output was processed through vep2sql.js:
"k8-linux bgt/misc/vep2sql.js -f 1kg11-1M.raw.vcf.vep > 1kg11-1M.raw.vcf.vep.fmf"

Running "bgt view -t CHROM,POS,REF,ALT -a,11:607967:0:CTCCGA 1kg11-1M.bgt" (the allele found in the 1kg11-1M.raw.vcf.vep.fmf file) returns nothing.

However, running "bgt view -t CHROM,POS,REF,ALT -a,11:608013:1:A 1kg11-1M.bgt" (the next allele on the list) does work.

Finding the insertion via searching the region does function and finds the variant.

bgt view -t CHROM,POS,REF,ALT -r 11:607960-607970 1kg11-1M.bgt

Running BCFTOOLS and BGT for the same region, BCFTOOLS showed two SNPs but BGT (bgt built from the same vcd.gz) returned none. Not sure how widespread this problem would be. Could someone take a look at this?

Examples like following:

bcftools view -r 05:67948215-67948219 xxx.vcf.gz | grep -v "#" | wc -l
2
bgt view -r 05:67948215-67948219 -s sample.list xxx.bgt | grep -v "#" | wc -l
0

sample.list includes all sample names from the VCF header.

Dear Li,

Since we have variant annotation (-d variantannot.fmf.gz option) and also the -a for query where we can select variants based on these annotations, it would be nice to have option(s) to include these annotations in the output (INFO field for VCF and also in the TAB format).

For example if we annotate variants by gnomAD frequency, we can query for it by the given rules, however when we have more complex queries we only know that the output comply to these variant selection criteria, however we don't know the exact values from the output. As the FMF syntax already have most the information (the type, the name, and also we know we have 1 per variant entry, we only lack a sensible Description however the name are usually self descriptive so it could be used. Guess it would be an overkill to have a separate table for descriptions for each FMF ID) to auto generate INPUT tags it would be nice if there would be an option to
#1) include variant annotations that were used in the -a query for filtering
#2) option to include all variant annotations (regardless of query)
#3) option to include a preset of selected annotations (regardless of query)

Since we have TAB format, it would be also nice to be able to reference these fields there too ie the -t option would work the fixed POS,CHROM, etc, plus these variant annotation fields (so we could list IMPACT, GENE, what else) also in the TAB format.

Zoltan

Just opening the bgt-server without queries we get INFO on config parameters and the queryable sample/variant annotations from the FMF files: XXX.bgt.spl for samples, and from the annot.XXX.fmf.gz that was provided by the -d option).

However there is no function to see what are the range in INT, or FLOAT fields or what are the queryable values in case of STRING fields.

Technically we usually know it or could document it based on our databses, however It would be nice to have some function to query these on the fly using the bgt fmf command (and/or with some new flags in the bgt-server). This would allow that users who have no access to the server could explore their options.

Regards,
Zoltan

I'm trying to get variants seen in only one sample X in the cohort.

I'm trying:

bgt view -s,X -f 'AC1>0&&AC<=2' -G cohort.bgt

It's not working. How can I achieve this?

Would be good to be able to reference files in queries. e.g. in bcftools, one can do something like

ID=@/path/to/file .. select lines with ID present in the file
ID!=@/path/to/file .. skip lines with ID present in the file

I guess I'm a little too quick off the mark, but is something missing from the Makefile?

make: *** No rule to make target bgtview.o', needed by bgt'. Stop.

Also, extracting k8 as in the instructions overwrites the bgt README.md with the k8 README.md

Hello.
As a test I ran BGT on chrX of 1KGP3 (available from the FTP link below) ftp://ftp.1000genomes.ebi.ac.uk/vol1/ftp/release/20130502/ALL.chrX.phase3_shapeit2_mvncall_integrated_v1c.20130502.genotypes.vcf.gz

The commands I used were (I converted the vcf.gz file above to BCF with bcftools to save space) :

bgt import chrX.bcf out/chrX
bgt view -b out/chrX.bgt > out/chrX.bcf

When I compared the output file with the input file with bcftools view, the first variant is strangely split.

Expected (only first sample GT shown because of size...) :

X       60020   .       T       TA,TAAC 100     PASS    AC=10,92;AF=0.00199681,0.0183706;AN=5008;NS=2504;DP=11848;AMR_AF=0.0029,0.0086;AFR_AF=0.0008,0.0635;EUR_AF=0,0.002;SAS_AF=0.0031,0;EAS_AF=0.004,0;VT=INDEL;MULTI_ALLELIC        GT      0|0 ...

And I got two variants :

X       60020   .       T       TA,<M>   0     .      .        GT      0/0 ...
X       60020   .       T       TAAC,<M> 0     .      .        GT      0/0 ...

I understand the the multi-allelic site has been split however what is strange is that in both lines I get genotype values between 0, 1, and 2.

So how to interpret when for example 2/0 occurs in the first line or in the second line ?

Thanks.
Best
Rick

Is there any way to import a VCF containing additional samples into an existing bgt file?

I've tried bgt import -S prefix.bgt 1.vcf followed by bgt import -S prefix.bgt 2.vcf, but this only seems to overwrite the existing file.

I cannot get freebayes vcf to import on 1.0-r282 at all

I get bgt: atomic.c:111: bcf_atomize: Assertion `i < b->n_info' failed when using bcfs that generated by freebayes --report-monomorphic and a seg fault on default vcfs of reporting only polymorphic (non-reference) sites.

In the release version of 1.0-r265, it works but not for monomorphic

The default freebayes of only polymorphic sites works in this version, but once again I get bgt: atomic.c:111: bcf_atomize: Assertion `i < b->n_info' using --report-monomorphic.

Let me know if you want backtraces, variable calls, or example files.