[lh3]

This is a flaw in bwa when generating SAM. I will fixed it.

It is not so easy to generate absolutely correct SAM due to the dependency between fields and between mates. We tried to minimize the dependency in design, but reducing dependency causes inconvenience in other cases. There is always a balance.

[corthay]

Thank you for your speedy response.

I have one more question. I got following results by using bwa(0.4.9), my favorite.
seq-name#0 69 * 0 0 * * 0 0 (sequence) (quality)
seq-name#0 133 * 0 0 * * 0 0 (sequence) (quality)

Both reads do not be mapped but the flag for "the mate is unnmapped" are 0.
How should I interpret it?

[jess]

Thanks Heng. I will try and let you know if I get stuck in something.

[lh3]

samtools.pl now updated at SVN:

http://samtools.svn.sourceforge.net/viewvc/samtools/trunk/samtools/misc/samtools.pl?view=markup

pileup2fq is implemented, similar to maq's cns2fq. Please note that samtools.pl filters based on the RMS mapping quality (-Q) while maq's cns2fq filters on the maximum mapping quality. Also, pileup2fq masks a small region around an potential indel, but maq's cns2fq does not. The overall accuracy looks similar to maq, though.

[jess]

ok! So i knw where i ws gng wrong...
the .aln file shud be put in last after all d options.

[jess]

I tried using the -c option,bt the pileup output is same evn widout this option! I gave d command smfink like dis:


samtools pileup -f ref.fasta aln_sorted.bam -s -c -v >test.pileup

Let me know wher I m gng Wrong!

[mikyatope]

Is there a way to convert a SAM consensus output (using -c option for pileup) to the old maq-style .cns consensus?

I have some maq-based pipelines I would like to use on my BWA results.

maybe it's related.

Is possible get the consensus sequence in a simple fasta format with SAMtools?

[jkbonfield]

I'm not sure if it's the case here, but I've noticed the CIGAR string has major issues if you attempt to include gaps in the clipped sequence.

Or rather CIGAR works fine I assume, but samtools does not. (It's not really a big issue as the only time I've seen this happen is someone manually trimming an alignment back.)

[lh3]

@corthay

You can convert with "blast2sam.pl -s" to save the sequence in SAM. Currently, samtools cannot parse SAM without sequence, although the specification allows this.

@jess and zee

Sorry. I am afraid that you cannot easily convert pileup output to maq-cns. You can call SNPs/indels directly from samtools and in SVN samtools.pl now filters SNPs/indels in the maq-like way. .cns has been replaced with glf, although they contain different information.

[corthay]

I would like to generate bam file from blast result of EST.

Firstly, I made sam file from blast result using blsat2sam.pl in samtools packages.
Secondly, I have tried to convert from sam file into bam file using "samtools import" but I've got following error.

Parse error at 1: CIGAR and sequence length are inconsistent.

seq1 16 chr20 18289430 255 23H426M760H * 0 0 * * AS:i:844 EV:Z:0.0

Could you give me any idea please ?

[jess]

Is there a way to convert a SAM consensus output (using -c option for pileup) to the old maq-style .cns consensus?

I have some maq-based pipelines I would like to use on my BWA results.

Yes, I am looking for the same. Heng li cud u help out plz?

[lh3]

X? tags are for aligner specific information. You need to read the documentation of the program that generates these tags.

[pparg]

I read through the latest SAM Format Specification on optional fields. But I still don’t understand some cases in the .sam file I got. For instances:
XT:A:R NM:i:0 X0:i:30406 XM:i:0 XO:i:0 XG:i:0
XT:A:R NM:i:1 X0:i:2 X1:i:1 XM:i:1 XO:i:0 XG:i:0 MD:Z:3G32
I just can’t figure out what do those X? fileds mean? I know these are are reserved for local use and re not be formally defined. But what is the point here? What are the types A, i, Z etc for?
Any hint/ideas would be appreciated.

[zee]

Is there a way to convert a SAM consensus output (using -c option for pileup) to the old maq-style .cns consensus?

I have some maq-based pipelines I would like to use on my BWA results.

[lh3]

On behalf of the Java API developers: the Java APIs and command-line tools are formally released here (also linked from http://samtools.sourceforge.net):

http://picard.sourceforge.net/

The Java implementation is more flexible than SAMtools-C on sorting/viewing/merging/rmdup. Developers may also find Java APIs are easier to use than the C APIs.

By the way, Lincoln Stein has implemented initial Perl bindings for the SAM/BAM format, based on the C APIs. This Perl module will become part of BioPerl as well as GBrowse.