stefpiatek / cnv-patissier Goto Github PK

In the unlikely event that a run is completed and then samples have been removed, this could lead to incorrect results (known cnvs or called cnvs). Remove a run's cnvs from these tables. Need to also remove all samples per run at the start of a rerun.

https://github.com/Illumina/canvas

https://github.com/theodorc/atlas-cnv

• Make sure a sample has only been used as normal_panel or as an unknown
• Make sure a sample hasn't been used as both normal and positive
• Make sure that the bed file matches the chromosome prefix type
• fasta.fai exists
• fasta.dict exists

https://github.com/girirajanlab/CN_Learn

Only really going to be used at GOSH because I can be certain of equal use of resources on the machine

Add md5sum collation to scripts and store in database

could be worthwhile to look at?

As a last step per cnv caller workflow, delete all other runs that aren't the last successful run from the config.

container did not start before specified timeout

maybe put a wait of a second between samtools view calls?

GATK 4.1 now used. gCNV out of beta but at the moment no best practices. Before sharing with other labs, check to see if there is best practices listed on GATK's website.

• requirements, installation, set up, usage, testing

https://github.com/rdemolgen/SavvySuite