Entering edit mode
4.9 years ago
Hello,
I got a local instance of galaxy on my computer, and wanted to run a differential expression analysis for some samples of human RNA-seq data. I've managed to download the fastq reads for the samples and I've downloaded and indexed the human genome using HISAT2 index builder. But when I tried to align the reads from just one of the six files, it wasn't able to align them in almost 60 hours of running. So I decided that I'd need to do this on the cloud. However, I don't want to have to reinstall the tools, and re-download all the files and re-index the genome. I was wondering if there was any way to take whatever I have on my computer, and put it onto the cloud as is, so that I can pick up where I left off.
Thank You
Please ask over at the Galaxy help forum for prompt response.
What is the command line you're using for hisat? Also, moving the indices, references, and tools into AWS is a considerable effort. If you're looking to just run a one-off analysis, you're better off launching a cloud instance, SSH-ing into the instance and performing your analysis there than setting up a Cloudman server, a Galaxy AMI, and a set of EBS volumes that automounts properly on your compute instances. The default cloudman will launch instances for you but it is still up to you to provide EBS volumes and troubleshoot any mounting problems you'll encounter. I think you're better off ssh-ing onto a compute node directly and running the analysis there unless you're trying to build infrastructure for a team, and you've given no indication that's your goal.
Alternatively, is there a cluster in your company/uni you could use? What are your parallelization options in terms of hardware you do have access to? How big are the fastq.gz files?