parse json to csv

[fanyucai1]

hi Developer,

I am bioinformatic from illumina in China.

So many dragen-customers(E.g:The First Affiliated Hospital of Zhengzhou University, China Yixueke Xueyuan Hemopathy Hospital) try to use our Nirvana software. But only csv files like annovar not json and fewer informations was needed.

1. the only one canonical transcript(RefSeq genes, not ENSEMBL) in clinical area

2. identify variants in specific genomic regions, identify variants that are documented in specific databases, for example, whether a variant is reported in dbSNP,1000 Genome Project, Exome Aggregation Consortium (ExAC) and Genome Aggregation Database (gnomAD)

3. calculate the SIFT/PolyPhen/LRT/MutationTaster/MutationAssessor/FATHMM/MetaSVM/MetaLR scores, find intergenic variants with GERP++ score<2 or CADD>10, or many other annotations on specific mutations.

4. annotation in clinvar,cosmic,intervar and omim

Thanks.

[MichaelStromberg]

Hi there,

This is something that we have attempted before, but unfortunately much of our data is hierarchical in nature and it's difficult to represent in a CSV or TSV file without losing some context in the process. In the early days of Nirvana, we did output the data into a TSV file, but the amount of data represented on each line became too much. The JSON files are much friendlier in larger sequencing analysis pipelines.

One of our developers, Haochen, recently created two Jupyter notebooks (one in Python and one in R) that makes parsing the Nirvana JSON files very easy. This way you can easily select the annotations that you are interested in filtering.

Please take a look at this page for more details and links to the Jupyter notebooks: https://illumina.github.io/NirvanaDocumentation/introduction/parsing-json

Best regards,

// Michael