Can I fix the human reference genome file on physical random access memory so that the bwa program can process it from this fixed file? Can a secondary bwa execution also align using this file without ram getting a new file? my goal is to ensure the amount of time and low ram usage during writing on random acces memory
The only aligner I'm aware of that puts its index into shared memory like that is STAR, so no there is not currently a way to do this with BWA. Note that BWA is quite memory efficient, so unless you have a node with 100+ threads available but only say 20GB RAM you shouldn't have issues with multiple BWA processes swamping your memory.