[axiom7]

Hi,

samtools-0.1.6_x86_64-linux; precompiled version downloaded today

$ bowtie --version
bowtie version 0.11.3
64-bit
Built on myserver
Fri Oct 23 13:27:05 MDT 2009
Compiler: gcc version 4.1.2 20070115 (prerelease) (SUSE Linux)
Options: -O3
Sizeof {int, long, long long, void*, size_t, off_t}: {4, 8, 8, 8, 8, 8}

$ samtools faidx hs_ref_chr10.fa
$ cat hs_ref_chr10.fa.fai
gi|89161187|ref|NC_000010.9|NC_000010 135374737 105 70 71

Created sam format with bowtie in two ways (behavior described below is the same for both methods):
1. using -S in bowtie command
2. using samtools bowtie2sam.pl on a bowtie --refout map file

sam file from method 2 (chromosome 10 only):

$ cat ref00000.map.sam
@0-0-3-9833 0 gi|89161187|ref|NC_000010.9|NC_000010 62380535 0 15M * 0 0 GCAAAGGNNATCATT IIIIIIIIIIIIIII NM:i:1 X1:i:5 MD:Z:3A3N0N6
@0-0-3-9833 16 gi|89161187|ref|NC_000010.9|NC_000010 62382480 0 15M * 0 0 GGGCTANNGCTCATC IIIIIIIIIIIIIII NM:i:1 X1:i:5 MD:Z:7N0N6
@0-0-6-12817 0 gi|89161187|ref|NC_000010.9|NC_000010 6095909 0 15M * 0 0 TACCACCNNGCCCTT IIIIIIIIIIIIIII NM:i:1 X1:i:265 MD:Z:1A5N0N6
@0-0-6-12817 16 gi|89161187|ref|NC_000010.9|NC_000010 6097174 0 15M * 0 0 GCATCANNCTCCCGA IIIIIIIIIIIIIII NM:i:1 X1:i:265 MD:Z:7N0N6


$ samtools view -bt ~/work/hs_ref_chr/hs_ref_chr10.fa.fai -o out.bam ref00000.map.sam
[sam_header_read2] 1 sequences loaded.
[sam_read1] reference '16 gi|89161187|ref|NC_000010.9|NC_000010 6097174 0 15M * 0 0 GCATCANNCTCCCGA IIIIIIIIIIIIIII NM:i:1 X1:i:265MD:Z:7N0N6

' is recognized as '*'.
[main_samview] truncated file.

out.bam is created, but I cannot do anything further with it.

Thanks.
Susan

[lsn]

SAM to MAQ converter

Dear All,
it might seem a strange question, but since many tools for detecting structural variants, such as BreakDancer and GASV, rely on a MAQ alignment file, I was wondering if anyone of you has already written a SAM to MAQ file converter.

Thank you in advance.

[nilshomer]

I prefer Maq's simulate over wgsim because, as far as I understand, wgsim generates errors in the reads uniformly, whereas Maq uses the quality scores of a set of actual reads that must be provided to Maq as a training set. Therefore, in addition to being less realistic, wgsim produces output that actually creates an inherent disadvantage for certain aligners (e.g. bowtie) that assume (usually correctly) that errors are not uniformly distributed, but that their incidence rises from the first base in the read to the last. Consequently, using wgsim as a means of generating simulated reads for alignment software comparison will not produce results that are as accurate as would be obtained by using Maq's simulate.

This brings me back to my original question: Is there a way to convert from SAM to Maq's .map?

FYI, the modified version of wgsim in DNAA uses per base error rates that can be input to the program.

In wgsim in samtools does not fairly represent color space data, since it is made for BWA. In short, BWA ignores the adapter and first base in color space. When wgsim simulates a color space read, it does not give the read with an adapter and then "n" color calls, but instead it creates an "n+1" length read, trims the adapter and first color, and gives a "n" length color read, so that BWA does not need to trim the adapter and first color. Therefore when user specifies a 50bp color space read, of which BWA should only be able to use 49bp, it outputs it so that BWA can use all 50.

I adapted wgsim into one of my own projects (http://dnaa.sourceforge.net) to correct this problem, as well as to make some of my own customizations, including the ability to give per base error rates instead of a uniform error rate.

Nils

[jmj1091]

Ah, I didn't even think to use wgsim_eval.pl; it hadn't occurred to me that I could use it without using wgsim. I just ran: maq2sam-long simu_align.map | perl wgsim_eval.pl. It worked perfectly, and gave me the same results as maq simustat simu_align.map.

Thank you very much!

[lh3]

You may simulate reads with "maq simulate" and evaluate the accuracy with "wgsim_eval.pl" given the SAM format. I think wgsim_eval.pl works with reads simulated with maq. Even if I am wrong, you may modify wgsim_eval.pl. There should not be much changes and it is a simple Perl script.

[jmj1091]

I prefer Maq's simulate over wgsim because, as far as I understand, wgsim generates errors in the reads uniformly, whereas Maq uses the quality scores of a set of actual reads that must be provided to Maq as a training set. Therefore, in addition to being less realistic, wgsim produces output that actually creates an inherent disadvantage for certain aligners (e.g. bowtie) that assume (usually correctly) that errors are not uniformly distributed, but that their incidence rises from the first base in the read to the last. Consequently, using wgsim as a means of generating simulated reads for alignment software comparison will not produce results that are as accurate as would be obtained by using Maq's simulate.

This brings me back to my original question: Is there a way to convert from SAM to Maq's .map?

[polivares]

You can use wgsim from SAM tools which as far as I understand is the "son" of mac simulate.

[jmj1091]

Hi,
I'm wondering if there's a way to convert from SAM to Maq's map format. I'd like to do this because I want to use Maq's simulate command to generate a set of simulated reads, align them (using an aligner other than Maq), output to SAM, then convert the SAM file to .map for use with Maq's simustat command. This way, I can get statistics for how well the other aligner aligned Maq's simulated reads. Does such a tool exist to make this possible?
Thanks!

[polivares]

@jasonbcold
I have not used jbrowse but I'd recommend you to try IGV which can load indexed BAM alignments without any trouble. If you need some helo PM me.

[jasonbcold]

Sam/bam to jbrowse

Hi there,
does anyone know, how to use/upload alignments in sam/bam format to jbrowse?
thanks,
peter

[houhuabin]

Hi heng,
I used samtools to built my bam file, but it dose't have a @HD field, which is needed by GATK's check for sort. How can I fix it?

[wumblehoots]

Pileup Base Quality

Hi, I was wondering if anyone could point me in the direction of any documentation that describes how to interpret the base quality information reported within a pileup format file.

I've checked here: http://samtools.sourceforge.net/pileup.shtml but can't find any information.

Cheers

[polivares]

http://seqanswers.com/forums/showthread.php?t=806

[lh3]

Do you have a small sample file for me to test? Thanks.

[torrey]

I was curious if anyone has found a workaround to this error? It is coming up for me when viewing novoalign alignments. I am using samtools 0.1.6 (r453), and get the following error when loading a samtools converted bam file from novoalign's sam file

samtools: bam_lpileup.c:116: tview_func: Assertion `l == tv->n_pre' failed.


thanks

I was wondering if I could get some help on the SAM specification.

I get the error:



When I page the viewer to my reads containing indels. A sample of my data is



And I'm using novoalign format where I've modified the novo2sam.pl script to produce a 'correct' CIGAR format. Most samtools operation work except for the tview