[gntc]

repeats

Depends whether the index was made from the masked version of hg19 or not. I'm pretty sure the pre-made index from the Bowtie website is made from the non-masked genome. Both masked and non-masked are available here:

http://hgdownload.cse.ucsc.edu/goldenPath/hg19/bigZips/

"chromFa.tar.gz - The assembly sequence in one file per chromosome.
Repeats from RepeatMasker and Tandem Repeats Finder (with period
of 12 or less) are shown in lower case; non-repeating sequence is
shown in upper case.

chromFaMasked.tar.gz - The assembly sequence in one file per chromosome.
Repeats are masked by capital Ns; non-repeating sequence is shown in
upper case."

The files in chromFa.tar.gz each start out with a large number of 'N's. Is this due to uncertainty near the ends of chromosomes in sequencing?

[ewels]

The files in chromFa.tar.gz each start out with a large number of 'N's. Is this due to uncertainty near the ends of chromosomes in sequencing?

What good timing - I was just searching out the answer to this exact question for that exact file... Does anyone know the answer?

[biznatch]

The files in chromFa.tar.gz each start out with a large number of 'N's. Is this due to uncertainty near the ends of chromosomes in sequencing?

I think yes, it is because of uncertainty near the ends of the chromosomes. If you look at hg19 in the UCSC Genome Browser and turn on the Gap track you can see where there are gaps in sequencing on each chromosome. Anywhere there is a gap will be N's in the sequence. There are gaps at the ends of each chromosome because telomeres and subtelomeres are repetitive and difficult to sequence and assemble. There are also large gaps at the centromeres for the same reason.

[droog_22]

Counting Hits in a BAM file

Dear All,

I am using bowtie to align reads to the dm3 genome. I just read that the SAM specifications allow for tags such as H0, H1, etc. which counts the number of 0-differences, 1-difference hits, and so on. I know how to do ass these tags using awk, I was just wondering if it would be straightforward to modify bowtie so that it outputs these values.

Cheers D.

[gntc]

mismatches

Dear All,
I am using bowtie to align reads to the dm3 genome. I just read that the SAM specifications allow for tags such as H0, H1, etc. which counts the number of 0-differences, 1-difference hits, and so on. I know how to do ass these tags using awk, I was just wondering if it would be straightforward to modify bowtie so that it outputs these values.

Bowtie does this automatically. The tag is XA:i:0 (for a read with 0 mismatches).

[kerhard]

limit to bowtie-build fasta input files?

Hi all,

I've been trying to make a bowtie index using a long list of annotated transposons as the input fasta files rather than reference chromosome files and bowtie-build does not seem to like it very much.

If I try to use ALL of the fasta files (which is a lot, probably around ~1000), I get the error message:

Error: could not open <fileX.fa>

But if I use only a subset of the fasta files (including fileX.fa), it works just fine.

I'm assuming that it's a memory issue, but the total contents of all of these fasta files is much less than the fasta files containing the full reference genome sequences, and I can make an index with them just fine.

Has anyone had any experience doing something similar? Is there some limit to the number of input files bowtie-build can take? I imagine that I can just split these files up into smaller groups and make several index files, but it would be nice to be able to have all of them in one index.

Thanks for any help/advice!

[kerhard]

bowtie-index can use single fasta file with multiple entries as input

woops, didn't realize that the bowtie indexer could take a single fasta file with multiple entries as input. seems to work just fine that way. because the bowtie website describes the input for bowtie-index as:

"A comma-separated list of FASTA files containing the reference sequences to be aligned to"

i assumed that they had to be separate files, but it looks like they don't.

sorry for the silly question, should have tried this simple solution first.

[dara]

bowtie -v option not working in version 0.12.3?

Hi all,

I'm finding an issue/possible bug or error on my part with one of bowtie's options, so I thought I would ask here.

I'm trying to use bowtie -v <int> option for a colorspace alignment and everytime I do, it takes me back to the usage parameters, indicating that something is wrong with the arguments.

I know that this works:

bowtie -C -f -n 3 -S -t hg18.cs.bowtie reads.csfasta aln.sam

But one I add the -v option, it doesnt, though it seems from the manual that it should:

bowtie -C -v 4 -f -S -t hg18.cs.bowtie reads.csfasta aln.sam

Any help/suggestions would be really appreciated.

thank you!

[azer]

-v 4 may be wrong. (1-3) is ok

[Gators]

So I am having some weird issues building a bowtie index with the hairpin.fa file from mirbase. The file was filtered to get rid of non-human miRNAs and adjusted to get rid of spaces. But I had the same problem before this filtering, etc. was done. I built the index with all default parameters. There is no obvious error message during the building procedure, but I am not sure I would catch anything unless there was the word "error." Anyway after building the index if I align with bowtie it reports back some alignments, but not even close to all of them. The vast majority of what it reports back have 1 or 2 mismatches (v was set to 2),although there are some perfect matches there. The sequences it reports back are also GC rich, which is weird to me...I also tried to build this on another computer and it gave similar results. So clearly something weird with the fasta file...

Any ideas?

I should say that other indexes have been built on this machine w/o problem, as has other alignments...

[BioSlayer]

bowtie never finishes nor read all ebwt indices

I have been following this thread from the beginning, I have few issues... the past four days I had a running bowtie instant and it never finished, had to kill it, my command was as follows

bowtie hg19 -q /CombinedReads/SRR065070_Combined.fastq -S  align.map --offrate 20 -p 2

So, looking around for possible causes I saw that an issue was registered at Sourceforge but was not followed up, I don't know if my situation in here is replicable but here are the factors that may have had some influence on that above behavior:
01- I downloaded indices from the bowtie website and unzipped that to a directory, that is the same directory I navigated to and ran the bowtie command from. So, I suppose bowtie could automatically relate to this. I took this measure since unzipping to the index folder within bowtie could not get it to read the indices (it kept complaining that it could not find an index hg19) so I created a directory and invoked bowtie from within it.
02- The file I get the reads from is downloaded from SRA (SRR065070), it is located in another directory from where I am calling bowtie (it is about 6 GBs) and has around 19 million reads. I used samtools to create the forward and backward reads in fastq format...
03- My system is a Ubuntu, 32 bits, 2 GB RAM, 7 GB SWAP.
04- The $bowtie --version output is


bowtie version 0.12.7
32-bit
Built on bio-laptop
Thu Apr 21 21:12:27 AST 2011
Compiler: gcc version 4.4.3 (Ubuntu 4.4.3-4ubuntu5)
Options: -O3 -Wl,--hash-style=both
Sizeof {int, long, long long, void*, size_t, off_t}: {4, 4, 8, 4, 4, 8}



Then, trying to investigate the behavior somewhat deeper (--verbose), I notice that out of the 6 ebwt indices (hg19.1.ebwt...hg19.4.ebwt, hg19.1.rev.ebwt and hg19.2.rev.ebwt), only four are being read (it just doesn't open hg19.3.ebwt nor hg19.4.ebwt), I tested that by passing a query from the STDIN...

$ bowtie hg19 -c acgggtttaa  test.map -t --verbose

and following is a brief excerpt from the output log
Opening hit output file: 15:16:21
About to initialize fw Ebwt: 15:16:21
About to open input files: 15:16:21
Opening "hg19.1.ebwt"
Opening "hg19.2.ebwt"
Finished opening input files: 15:16:21

About to initialize rev Ebwt: 15:16:21
About to open input files: 15:16:21
Opening "hg19.rev.1.ebwt"
Opening "hg19.rev.2.ebwt"
Finished opening input files: 15:16:21
Reading header: 15:16:21

About to open input files: 15:16:21
Opening "hg19.1.ebwt"
Opening "hg19.2.ebwt"
Finished opening input files: 15:16:21
Reading header: 15:16:21

About to open input files: 15:16:44
Opening "hg19.rev.1.ebwt"
Opening "hg19.rev.2.ebwt"
Finished opening input files: 15:16:44


Seeking your guidance and support with appreciation...

[harrike]

Here maybe the best place for my question.

I am trying to use Bowtie to map about 300,000 reads to my reference. I use the command: bowtie -a -v 0 -p 10 -t INDEX_FILE -f READS_FILE.fasta > RESULT_FILE --un unmapped.txt.

The command couldn't be finished, an error showed up "You Mac OSX startup disk has no more space available for application memory". The Bowtie process took about 30 GB memory and froze.

I checked my startup disk (Macintosh HD). It still has about 1 Tb space available. I don't know what's going on.

I am using a computer with 16 cores and 32 GB memory.

Hope somebody here can help me. Thanks in advance.

[sdvie]

Here maybe the best place for my question.

I am trying to use Bowtie to map about 300,000 reads to my reference. I use the command: bowtie -a -v 0 -p 10 -t INDEX_FILE -f READS_FILE.fasta > RESULT_FILE --un unmapped.txt.

The command couldn't be finished, an error showed up "You Mac OSX startup disk has no more space available for application memory". The Bowtie process took about 30 GB memory and froze.

I checked my startup disk (Macintosh HD). It still has about 1 Tb space available. I don't know what's going on.

I am using a computer with 16 cores and 32 GB memory.

Hope somebody here can help me. Thanks in advance.

are you executing this command from your Mac on a remote machine with 16 cores and 32 GB, or does your Mac have 16 cores and 32 GB itself ?

The TB space available will not help if the application freezes because of lack of RAM.

[harrike]

Thanks, Sdvie.

It my Mac which has 16 cores and 32 GB. I don't use remote control.

[sdvie]

Thanks, Sdvie.

It my Mac which has 16 cores and 32 GB. I don't use remote control.

I am suprised, that bowtie seems to be so memory-intensive in your case... especially with relatively few reads. Did you use a particularly large genome?

cheers,
Sophia