Tried (but no better results)
- Different exome analysis CNV tools: XHMM, ExomeCopy, cn.mops, Conifer
- PennCNV (SNP array CNV)
- Filtering CNVs based on length, CNV type (dup/del), origin (denovo, inherited, all), number of markers, gene-coding regions, brain-expressed regions, literature-based search, removing well-known variants
Still to try
- Play with tool-specific tuning parameters (e.g. HMM priors, variant threshold sensitivities)
- Dive deeper into gene-based analysis (e.g. permutation analysis for combination of genes)
- Contrast with multiplex samples (waiting on Tri + Sijung new pipeline for final results)
- Check if worth doing anything with imaging for the 25 kids.
- Add more phenotypes for comparison, and more pipelines in the Y axis