Stream: genomics
Topic: VCF2FHIR: Exomiser results
Marta Herranz (Oct 20 2020 at 15:54):
Hi! I am doing a genetic data mapping and would like to know if anyone has worked with this before.
After running exomiser (https://exomiser.github.io/Exomiser/) with a set of symptoms and VCF files I get the gene information for a particular case. In this result I have data from:
- Gene identifier
- Mode of inheritance
- Some scores: combined, phenotype and variant.
- Contributing variants: chromosome, position, ref and alt, frequency data, variant effect, clinVar data (alleleId, interpretations, reviewStatus), etc.
From what I have read so far, the genetic information is stored in "Observation" resources: http://hl7.org/fhir/observation-genetic.html. And that this also has several extensions that I could use.
With all the observation resources created for a patient, then I would have to create a reporting resource that points to all of the above, is that right?
I have been investigating if I could do this mapping automatically since the result files are long enough to try to do it one by one, and I have found the following:
How can I save the genetic information in VCF format I can use this converter to go to FHIR: https://github.com/openelimu/VCF-2-FHIR
Has anyone ever used it or done this conversion from VCF to FHIR format more or less automatically?
Thank you very much!
Joel Schneider (Oct 20 2020 at 17:06):
The observation-genetics profiles are deprecated and will be removed in FHIR R5, so those are presently NOT recommended for use.
Instead, the workgroup recommends implementers to utilize the Genomics Reporting Implementation Guide:
http://hl7.org/fhir/uv/genomics-reporting/
Additionally, work on an updated revision to the Genomics Reporting Implementation Guide is currently in progress, and is expected to be ready for publication soon:
http://build.fhir.org/ig/HL7/genomics-reporting/
Bob Dolin (Oct 20 2020 at 17:18):
Hi @Marta Herranz . We've released an updated version of the vcf2fhir converter here [https://github.com/elimuinformatics/vcf2fhir]. We plan a future enhancement to include translation of additional annotations (see https://git.hub.com/elimuinformatics/vcf2fhir/issues/9).
Jamie Jones (Oct 20 2020 at 21:07):
I'll also add in that the chromosome field will be added in an update to http://build.fhir.org/ig/HL7/genomics-reporting/StructureDefinition-variant.html. We are currently upgrading internal tooling on our implementation guide but further developments will resume soon.
Bob Dolin (Oct 21 2020 at 00:27):
@Marta Herranz I've never used exomiser, but I have, for instance, run a VCF file through snpEff and snpSift, to add ClinVar and other annotations. If I understand what you're trying to, it would be, for instance, to take the annotated VCF, and automatically convert it into FHIR Genomics format?
Here for instance is a single row of a VCF containing annotations. I think it would be pretty straightforward to map this into FHIR format, although would need to make some design / mapping decisions.
chr1 2556714 rs4870;135349 A G 195.77 PASS AC=1;AF=0.500;AN=2;BaseQRankSum=-1.079e+00;DB;DP=22;ExcessHet=3.0103;FS=7.282;MLEAC=1;MLEAF=0.500;MQRankSum=0.00;POSITIVE_TRAIN_SITE;QD=8.90;ReadPosRankSum=0.569;SOR=0.452;VQSLOD=0.688;culprit=QD;ANN=G|missense_variant|MODERATE|TNFRSF14|ENSG00000157873|transcript|ENST00000355716.4|protein_coding|1/8|c.50A>G|p.Lys17Arg|349/1707|50/852|17/283||,G|missense_variant|MODERATE|TNFRSF14|ENSG00000157873|transcript|ENST00000435221.6|protein_coding|2/7|c.50A>G|p.Lys17Arg|581/1225|50/694|17/230||WARNING_TRANSCRIPT_INCOMPLETE,G|missense_variant|MODERATE|TNFRSF14|ENSG00000157873|transcript|ENST00000426449.5|protein_coding|2/6|c.50A>G|p.Lys17Arg|159/660|50/551|17/182||WARNING_TRANSCRIPT_INCOMPLETE,G|missense_variant|MODERATE|TNFRSF14|ENSG00000157873|transcript|ENST00000434817.5|protein_coding|2/7|c.50A>G|p.Lys17Arg|190/834|50/694|17/230||WARNING_TRANSCRIPT_INCOMPLETE,G|missense_variant|MODERATE|TNFRSF14|ENSG00000157873|transcript|ENST00000451778.5|protein_coding|2/7|c.50A>G|p.Lys17Arg|183/827|50/694|17/230||WARNING_TRANSCRIPT_INCOMPLETE,G|missense_variant|MODERATE|TNFRSF14|ENSG00000157873|transcript|ENST00000409119.5|protein_coding|2/7|c.50A>G|p.Lys17Arg|170/896|50/582|17/193||,G|upstream_gene_variant|MODIFIER|TNFRSF14-AS1|ENSG00000238164|transcript|ENST00000432521.2|pseudogene||n.-2564T>C|||||2564|,G|upstream_gene_variant|MODIFIER|TNFRSF14-AS1|ENSG00000238164|transcript|ENST00000449660.5|pseudogene||n.-4608T>C|||||4608|,G|upstream_gene_variant|MODIFIER|TNFRSF14|ENSG00000157873|transcript|ENST00000466750.5|retained_intron||n.-1099A>G|||||1099|,G|upstream_gene_variant|MODIFIER|TNFRSF14|ENSG00000157873|transcript|ENST00000463835.5|retained_intron||n.-3840A>G|||||3840|,G|upstream_gene_variant|MODIFIER|TNFRSF14-AS1|ENSG00000238164|transcript|ENST00000443892.2|pseudogene||n.-2469T>C|||||2469|,G|upstream_gene_variant|MODIFIER|TNFRSF14|ENSG00000157873|transcript|ENST00000471768.1|retained_intron||n.-2866A>G|||||2866|,G|upstream_gene_variant|MODIFIER|TNFRSF14|ENSG00000157873|transcript|ENST00000482602.5|protein_coding||c.-3969A>G|||||3968|WARNING_TRANSCRIPT_NO_START_CODON,G|intron_variant|MODIFIER|TNFRSF14|ENSG00000157873|transcript|ENST00000475523.5|retained_intron|1/5|n.70+259A>G||||||,G|intron_variant|MODIFIER|TNFRSF14-AS1|ENSG00000238164|transcript|ENST00000416860.2|pseudogene|1/5|n.36-18T>C||||||,G|intron_variant|MODIFIER|TNFRSF14-AS1|ENSG00000238164|transcript|ENST00000452793.1|pseudogene|1/3|n.56-18T>C||||||,G|non_coding_transcript_exon_variant|MODIFIER|TNFRSF14|ENSG00000157873|transcript|ENST00000442392.6|processed_transcript|1/4|n.310A>G||||||,G|non_coding_transcript_exon_variant|MODIFIER|TNFRSF14|ENSG00000157873|transcript|ENST00000482074.6|retained_intron|1/3|n.309A>G||||||,G|non_coding_transcript_exon_variant|MODIFIER|TNFRSF14|ENSG00000157873|transcript|ENST00000496064.5|retained_intron|1/5|n.158A>G||||||,G|non_coding_transcript_exon_variant|MODIFIER|TNFRSF14|ENSG00000157873|transcript|ENST00000463471.6|retained_intron|1/5|n.144A>G||||||;AF_ESP=0.55803;AF_EXAC=0.54431;AF_TGP=0.61482;ALLELEID=139088;CLNDISDB=MedGen:CN169374;CLNDN=not_specified;CLNHGVS=NC_000001.11:g.2556714A>G;CLNREVSTAT=no_assertion_provided;CLNSIG=not_provided;CLNVC=single_nucleotide_variant;CLNVCSO=SO:0001483;CLNVI=UniProtKB_(protein):Q92956#VAR_013007;GENEINFO=TNFRSF14:8764|TNFRSF14-AS1:115110;MC=SO:0001583|missense_variant;ORIGIN=1;RS=4870 GT:AD:DP:GQ:PL 0/1:13,9:22:99:224,0,394
Marta Herranz (Oct 22 2020 at 08:52):
Thank you all very much for your help.
I have tried to install the release of the converter from https://github.com/elimuinformatics/vcf2fhir but I am having problems.
I cloned the github project and followed the steps explained at https://vcf2fhir.readthedocs.io/en/latest/user/install.html but it won't let me run the "pip install vcf2fhir" command. It gives me errors with pysam:cython, I read and thought that first I had to run the "python setup.py build" and then the installation, but it still doesn't work. Besides, I can't use vcf2fhir/bin/activate source either because in the vcf2fhir folder when I clone the project there is no bin folder.
I have previously installed https://github.com/openelimu/VCF-2-FHIR, can it be that I am having problems with that?
Can someone tell me the steps to install it and be able to use it?
Bob Dolin (Oct 22 2020 at 13:59):
Hi @Marta Herranz I'll respond to your PM
Last updated: Apr 12 2022 at 19:14 UTC
