Entering edit mode
6.8 years ago
XBria
▴
90
Hi,
Last time I mapped my data with STAR that took me less than two minutes. But this time for 0.1 percent of data it took 45 minutes to map !!! The only differences are these : data size already mapped was 290 MB (on virtual box ubuntu). and this time it is about 800 MB (on dual boot ubuntu) Can anyone please let me know why there is such a huge difference in mapping time ?
In this case, If I increase the volume of data to 0.2 , 0,4 , .. it will take much more time to map ?! can I go further with a system of 8 RAM memory ?!
Thanks.
Lacking further details, I'm guessing that it's slow due to running out of RAM. I normally use 20 or so cores and map a dataset of <20GB in 10-15 minutes. It'd be faster if you have local storage, since we're doing a lot over NFS.
Dear Devon,
Why did the lack of RAM memory not affect the time factor in last mapping ? but this time, again the same system features, only the size of data is increased from 290 to 800 MB. which details do you need to know ? ................ I am mapping only on the chromosome X using STAR. Data are paired-end. 75 bp length. HDD 256
I was presuming that you had it sorting the BAM in memory. If you're just mapping to chrX then 8GB is probably sufficient, I don't then have any good explanation for the slow performance.
Can you do a quick check by mapping only around 300MB of current data and compare the time? BTW, also check that the command line that you used in last mapping is the same as this one (you can get the command line by STAR output files). If everything fails to give a reasonable answer, please post your command line here.
What is your RNA-seq material, and What is the reference genome are you using ?