Entering edit mode
8.6 years ago
lvogel
▴
30
Hi, I have to run tophat2 on the Linux command line, but my fastq reads are stored on another remote location, due to their very large size (even when gzipped). Is this possible, and how do I do it? Thanks. Also, I'm using Amazon Web Services, if this makes a difference.
If your Amazon cloud has SSH service, use SSHFS to mount the location from Amazon in your local server/computer.
1) sshfs USER@amazon.server:/home/REMOTEUSER/LOCATION/TO/FOLDER/HAVING/FASTQ /home/LOCALUSER/SOMEFOLDER/
After entering above command, it will ask for password
ENTER amazon ssh PASSWORD for you account
2) tophat2 /home/LOCALUSER/SOMEFOLDER/FASTQ_filename.gz
Have you actually done something like this? Is it worth the trouble in terms of performance? I am asking to get first-hand information.
"another remote location"as in outside AWS? If you are able to export (nfs or otherwise) and then "mount" that location on your amazon machine it may (in theory) be possible to compute on the data. It would not be very advisable in terms of performance/security (assuming you are able to make the exports available in the first place).Since you are asking how to do it I have a feeling you don't have the right expertise. Bite the bullet and transfer the data to AWS if that is the only place you can compute.
What I mean by "another remote location" is that I have tophat2 installed on one server in AWS, and the reads in another one that's meant for storage, because putting all the fastq reads all on the one for computing (EC2) was getting too expensive. So, both locations are remote from my local laptop, but both are AWS.
(Sorry if I wasted anyone's time by not including this information.)
Not an AWS expert but @EagleEye's solution should work. Others options may exist and people would chime in with those.
OK. I'll let you know how it works out.