BED file for FMI Target-Seq Regions

[edsu7]

As per https://github.com/icgc-argo/workflow-roadmap/issues/377:

We'll need to generate a bed file documentating regions highlighted in https://info.foundationmedicine.com/hubfs/FMI%20Labels/FoundationOne_CDx_Label_Technical_Info.pdf

[edsu7]

Data posted in google link: exons.tsv f1_exons.tsv f1lcdx_exons.tsv f1cdx_exons.tsv

• contains coordinates for exons per ensembl exon ID

exon_genes.tsv - contains ensembl gene ID, gene name, synonyms, ensembl transcript IDs

introns_and_other_elements.tsv f1_introns.tsv f1lcdx_introns.tsv f1cdx_introns.tsv

• contains element coordinates, affiliated gene transcript and genomic element type

intron_genes.tsv gene_mapping.tsv

• contains ensembl gene ID, gene name, synonyms, type of feature annotated targeted by FMI, introns targeted by FMI, transcripts, transcripts with introns, transcripts without introns.

Steps for producing files:

• use ensembl biomart and pull genes for GRCh38.p14
• matched FMI's gene list to ensembl IDs
• use ensembl's REST API to retrieve info associate to each gene
  • e.g. https://rest.ensembl.org/lookup/id/ENSG00000164362?expand=1
  • results from those queries are saved in intron_genes.tsv and exon_genes.tsv
• for exons, pull per transcript in payload and save into exons.tsv
• introns are harder
  • pull exons per transcript
  • subtract exons from transcript
  • pull appropriate intron according to FMI (and only if # of introns is present i.e. cannot pull intron #8 in only 3 introns present in transcript)
• for ncRNA use full gene coordinates
• for promoter take TSS and +/- 500 bp and +/-1000 bp depending on strand

[edsu7]

Complete. Closing.