kunalkathuria / SVXplorer

Structural Variant Caller
MIT License
9 stars 2 forks source link

Notes for run-time for sims and real-data #63

Closed kunalkathuria closed 4 years ago

kunalkathuria commented 6 years ago

Run with:

Good BED regions file, including GC-bias regions Sec almts, appropriately spaced if necessary Analyze breakpoints overlap only without svtype, and one such analysis for all INS pooled together.

kunalkathuria commented 6 years ago

Loop over some params like weight threshold

kunalkathuria commented 6 years ago

Need to consider whether to use exclude file for all tools or not for simulations (is it artificial?)

kunalkathuria commented 6 years ago

Assess efficacy of SPLIT_INS. See if losing anything unreasonable in Sim2.

kunalkathuria commented 6 years ago

MQ 1 vs 10? Sims vs real data?

kunalkathuria commented 6 years ago

What slop value is reasonable? 300 seems more reasonable (really one-sided if you think about it).

kunalkathuria commented 6 years ago

INS_C dels

kunalkathuria commented 6 years ago

Decide on Final Good Regions File or no file at all after assessing real data results

kunalkathuria commented 6 years ago

MQ_U = 10?