QBD is a system to Query By Data.

Requirements for the system

• genomics vcf
• transcriptomics counts
• proteomics counts
• epigenomics peaks

1. Find all patients with mutations in a set of genes {X}.

   Annotate the variants based on their frequency in a control population, likely impact on function, etc.

   Return the list of patients and iPSC lines

   (I imagine this would be a search box on the website, and expect it would be very valuable for people looking for iPSC lines -- it's related to your example I.3)

2. Run all the coding mutations through some or all of the following tools:

   • SIFT
   • PolyPhen
   • GERP++
   • Condel
   • CADD
   • fathmm
   • MutationTaster
   • MutationAssessor
   • GESPA
   • REVEL

   (This list was stolen from a review: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5433009/ )

   Generate a table of the results suitable for further processing

3. a. Find all non-coding variants in each patient that overlap with that patient's ATAC-Seq peaks

   b. Generate a consensus set of ATAC-Seq peaks and find the non-coding variants for all patients.

4. a. Use output from filters like those in cases 2 and 3 to generate a table of patients vs. variants.

   b. Cluster patients by similarity in variants

   c. Run dimensionality reduction algorithms on this table