TopHat approximate run time & memory usage?

xinchen · 05-16-2010, 08:04 PM

Hi everyone,

I'm using TopHat to map my RNA-seq reads to splice junctions for use in Cufflinks, but it's been taking a bit longer than I expected. One sample I ran (as an initial test) has been going for ~34 hours, and the tophat_out folder is using up around 76 GB of space. By looking in the logs, it seems to be on the "segment_juncs" step.

Each of the individual samples I hope to align have roughly 90 million reads (@ 50 nt) over 3 lanes, and I'm aligning to the human genome (hg19).

Would anyone know how long I should expect the program to run for, and also how much disk space it'll need per sample?

Thanks!

edit: I'm using single-end reads

shurjo · 05-16-2010, 08:20 PM

It routinely takes ~40 hours for me (~45 million reads, with - p 4 to use four threads and 16G RAM). It will delete the huge temp files it generates, so the disk usage may not be that much of any issue.

sphil · 05-17-2010, 12:23 AM

I'm not sure but i think that TopHat need for 75million reads about 3 Days...

xinchen · 05-17-2010, 02:03 PM

Thanks! It ended up taking ~50 hours for my test sample, which isn't too long

ymc · 05-18-2010, 03:47 AM

How can you guys compare run time and memory usage without stating the CPU and RAM you are using???

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Tophat Run time	bassu	Bioinformatics	11	09-17-2013 10:59 PM
Tophat memory usage during "Searching for junctions via segment mapping"	biznatch	RNA Sequencing	9	02-18-2013 10:47 AM
segment_juncs memory usage while running Tophat	genec	Bioinformatics	1	11-22-2011 08:09 AM
Memory usage	Seta	General	2	04-05-2011 11:44 AM
Memory Usage in Newbler 2.3	smg283	454 Pyrosequencing	5	11-09-2010 08:46 AM

05-16-2010, 08:04 PM	#1
xinchen Junior Member Location: Toronto Join Date: May 2010 Posts: 6	TopHat approximate run time & memory usage? Hi everyone, I'm using TopHat to map my RNA-seq reads to splice junctions for use in Cufflinks, but it's been taking a bit longer than I expected. One sample I ran (as an initial test) has been going for ~34 hours, and the tophat_out folder is using up around 76 GB of space. By looking in the logs, it seems to be on the "segment_juncs" step. Each of the individual samples I hope to align have roughly 90 million reads (@ 50 nt) over 3 lanes, and I'm aligning to the human genome (hg19). Would anyone know how long I should expect the program to run for, and also how much disk space it'll need per sample? Thanks! edit: I'm using single-end reads Last edited by xinchen; 05-16-2010 at 08:06 PM.

05-16-2010, 08:20 PM	#2
shurjo Senior Member Location: Rockville, MD Join Date: Jan 2009 Posts: 126	It routinely takes ~40 hours for me (~45 million reads, with - p 4 to use four threads and 16G RAM). It will delete the huge temp files it generates, so the disk usage may not be that much of any issue.

05-17-2010, 12:23 AM	#3
sphil Senior Member Location: Stuttgart, Germany Join Date: Apr 2010 Posts: 192	I'm not sure but i think that TopHat need for 75million reads about 3 Days...

05-17-2010, 02:03 PM	#4
xinchen Junior Member Location: Toronto Join Date: May 2010 Posts: 6	Thanks! It ended up taking ~50 hours for my test sample, which isn't too long

05-18-2010, 03:47 AM	#5
ymc Senior Member Location: Hong Kong Join Date: Mar 2010 Posts: 498	How can you guys compare run time and memory usage without stating the CPU and RAM you are using???