At the moment benchmark_consesnus.sh was launched for folder which is manually created and cleaned from outliers e.g. /mnt/stripe/bio/experiments/configs/benchmark/benchmark/H3K27me3/macs_broad/clean/. It will be nice if we can do such cleanup automatically, without human factor. So I see 2 ways:
1) Do some cleanup script, which takes info from our current outliers.csv
2) Pass outliers.csv to benchmark_consesnus.sh so as allow it ignore outliers
Also we have several tool related folders (zinbra, sicer, zinbra_noinput, macs2, ..), so we can pass this tool folder name to benchmark consensus script, instead of hardcoded folders
At the moment
benchmark_consesnus.sh
was launched for folder which is manually created and cleaned from outliers e.g./mnt/stripe/bio/experiments/configs/benchmark/benchmark/H3K27me3/macs_broad/clean/
. It will be nice if we can do such cleanup automatically, without human factor. So I see 2 ways: 1) Do some cleanup script, which takes info from our current outliers.csv 2) Pass outliers.csv tobenchmark_consesnus.sh
so as allow it ignore outliers