[simobioinfo]

Thank you everybody. I was on vacation. I will try soon indexes you sent me

[dpryan]

@simbioinfo: I just sent you a PM with some links to the bowtie2 indices that worked on my Mac. Let me know if they work for you.

[dpryan]

I was able to build the index on another machine and use it on my macbook pro with 3 gigs of RAM. I really wouldn't recommend doing this on a machine with that little RAM as you end up swapping all the time. Since you have more RAM than my old laptop, hopefully you'll get better performance. I'll post the indices later today and send you a link.

[simobioinfo]

How much RAM does this Mac have? If you look at my previous post you'll see where I said that if you don't have enough RAM it will take very, very long to build in the index. To build the human genome you probably need 5-6GB but you will have to monitor RAM usage as the index is building to know for sure. And it doesn't matter how many cores you have Bowtie2-build is a single process.

My mac has 4Gb of RAM. I suppose that it could be enough to build index. I don't know how to check ram usage.

[biocomputer]

Hi... now 24 hours are passed and indexing is still running on my MAc (see previous post of this thread). What can I do now? My processor is dual core... I have to insert an option into the command line to take into account that processor has 2 core?

How much RAM does this Mac have? If you look at my previous post you'll see where I said that if you don't have enough RAM it will take very, very long to build in the index. To build the human genome you probably need 5-6GB but you will have to monitor RAM usage as the index is building to know for sure. And it doesn't matter how many cores you have Bowtie2-build is a single process.

[dpryan]

I'll give it a whirl on my macbook pro, but it's rather old and low on RAM. Assuming that doesn't work, on Tuesday I'll be back in the office and just use one of our beefier Macs if this isn't resolved by then.

[GenoMax]

simobioinfo is indexing GRCh37/human. If bandwidth is not a problem then you can download the preformatted bundle (sequence, indexes and annotations) from this link: http://support.illumina.com/sequenci...e/igenome.html

@simobioinfo: This is the sort of thing I was trying to warn you about in terms of user experience.

@Devon: If you have indexes for that known to work on a Mac you could post them for simobioinfo.

[dpryan]

24 hours seems excessive. Can you see it still using one of your cores? I don't think the indexing can be run multi-core. What species are you trying to index? Perhaps one of us can just create and post the indices so you can download them.

[simobioinfo]

Hi... now 24 hours are passed and indexing is still running on my MAc (see previous post of this thread). What can I do now? My processor is dual core... I have to insert an option into the command line to take into account that processor has 2 core?

[dpryan]

Yeah, they'll take a while to get generated.

[simobioinfo]

Don't be surprised if it take a couple hours. For reference, on my (likely faster) machines it takes ~1.5 hours to index the human or mouse genome. You can follow the progress by looking at the sizes of the 6 files produced. I recall that the .3.bt2 and .4.bt2 files are produced first and the first half of the process is done once the .2.bt2 file is about the same size as the .4.bt2 file. The .rev.1.bt2 and .rev.2.bt2 files will be the same size as their forward counterparts, so you can gauge how far along they are once that half of the indexing commences.

Thank you. Indexing is started now. I have 4 files:
3.bt2 of 4kb
4.bt2 of 716.2 Mb
1.bt2 and 2.bt2 are 0 byte at the moment.
Is that ok?

[dpryan]

Don't be surprised if it take a couple hours. For reference, on my (likely faster) machines it takes ~1.5 hours to index the human or mouse genome. You can follow the progress by looking at the sizes of the 6 files produced. I recall that the .3.bt2 and .4.bt2 files are produced first and the first half of the process is done once the .2.bt2 file is about the same size as the .4.bt2 file. The .rev.1.bt2 and .rev.2.bt2 files will be the same size as their forward counterparts, so you can gauge how far along they are once that half of the indexing commences.

[simobioinfo]

I will try...

[dpryan]

You just want "bowtie2-build genome.fa genome", without the asterisk and redirect ('>').

[simobioinfo]

Hi again... first of all I would like to thank you everybody.
Yesterday I tried to index human genome on my Mac with this command:
bowtie2-build * genome.fa>genome
Shell has been showing me this line:"building SMALL index" for 4 hours.
Did I make some error launching indexing? or is it normal?
I don't know what to do now
Simona