Description
CIViC is an open access, open source,
community-driven web resource for Clinical Interpretation of Variants in
Cancer.
Variants are mapped to the genomic location, unless no RefSeq transcript
matches the reference amino acid of the variant. Gene-level data is mapped to
the location of the RefSeq transcript of the gene.
Display conventions
Each combination of (gene, variant, disease) tuple in the database is represented by an individual feature.
Mouseovers show the drugs. Clicking on the feature shows all details associated to the entry in CIViC.
This is very "sparse" track, with only 195 features and a total genome coverage
of 3Mbp. The ten most annotated genes in this track are the typical cancer
genes (number of database entries in parentheses): NPM1 (39), BRAF (30), DNMT3A
(29), KRAS (28), FLT3 (28), KIT (22), PIK3CA (19), WT1 (16), NRAS (14), EGFR (13).
The last field "Mapped to genome via" was added by the mapping pipeline and is
not part of the database. It shows, using HGVS, how the variant was mapped to
the genome. The mapping first checks the amino acid on all RefSeq Protein
sequences at the indicated position, maps this to the corresponding RefSeq cDNA
sequence and then uses pslMap to find the correct hg19 position.
Fusion genes are mapped to the gene locations of both genes of the fusion.
Data access
The full database can be downloaded from CIViC.
The nightly dump from Mar 1 2016 was used for this track.
The mapped locations on hg19 are available as a bigBed file.
Credits
Thanks to Obi and Malachi Griffith for support with the data import.
References
"Clinical cancer sequencing and integrated analysis of whole genomes, exomes and transcriptomes"
AACR Special Conference: Translation of the Cancer Genome, February 7-9, 2015
|
|