Replace RNA expression analysis by DROP pipeline

[marc-sturm]

germline - rare disease

• optimal cohort size
• cohort selection
  • no relatives
  • only indices?
• RNA QC
• exons/genes that are lowly/not expressed
• validation
  • check for RNA-WGS/WES/lrGS for causal variants (CNVs, splicing variants, ...)
  • trio de-novo > check RNA
• 2. analysis - with gene list based on HPO > multiple testing correction better
• expression on gene-level, splicing-analysis for transcript- and exon-level?!
• data sharing? per gene: variant type, expression, splicing-data
• alternative splicing example not properly detected by cancer
• examples of currently not detected events:
  • alternative splicing caused by chr20:1164481-1164481 G>A in 22100a030C3_01

======

• cohort selection
  • only tumor RNA of same tissue/cancer type
  • exclude samples with CNVs in genes with known tumor-specific CNVs (pre-defined list)