[jnfass]

samtools pileup for multiple diploid individuals?

Is it appropriate to use samtools pileup (which uses maq's consensus- and SNP-calling model) on pooled reads from multiple diploid individuals? I'm looking for SNPs both within the read population, and between the reads and the reference. I've got 6 individuals in separate samples from a species close to the reference, at low enough depth where I should probably just forget about mapping each individual separately (in other words, I can forget about calling a genotype for each individual). But I'd still like to call the most likely consensus and SNP for this population ...

[bekkari]

Does any one know about any tool/program to convert SAM output to BED format, if so please let me know

[zlu]

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.

Has anyone had any luck with this? In addition, in MAQ, you can dump unmapped reads into a separate file, is there such a function/tool in samtools? Thank you.

[apfejes]

bekkari:

I believe ConvertToBed.jar in the Vancouver Short Read Analysis Package can do it.

Anthony

[bekkari]

if you use BOWTIE as an alignment algorithm, there is an option (--un) to dump all unmapped reads into a file.

[zlu]

if you use BOWTIE as an alignment algorithm, there is an option (--un) to dump all unmapped reads into a file.

But I'm using BWA. Is there an unmapped flag in the sam file?

[xiang]

Partial pileup for samtools?

samtools pileup takes ages, so does varfilter.

Can samtools pileup work on one chromosome? It would be
must easier for parallelization.

[lh3]

http://samtools.sourceforge.net/pipe.shtml

[zlu]

properly mapped Flag

Perhaps I have misunderstood it but isn't right that properly mapped flag (P string) are only used when read pairs are mapped to the same chromosome with correct insert size? I have about 6% of the properly mapped reads with the P string flag that have mate mapped to different chromosome with 0 insert size as shown below. Has anyone seen this before?

EBRI093151:1:90:555:299#0 pPR1 Chr11 107308221 23 36M Chr12 49 0 TATCCTATTCGAAAGTCGCCATGACCGTGGACATGA BCCBCBBACCB?CBA@BBACCBCAAB<6<?BABBBB XT:A:U NM:i:0 SM:i:23 AM:i:23 X0:i:1 X1:i:1 XM:i:0 XO:i:0 XG:i:0 MD:Z:36

EBRI093151:1:90:555:299#0 pPr2 Chr12 49 23 11S7M1D12M6S Chr11 107308221 0 CTACCGCTTGGGTGGTCATGAATGATTAGCACGCCC AB99@B=BBA>ACCBCCBBCBBCBBBBCBCBB@B@A XT:A:M NM:i:4 SM:i:23 AM:i:23 XM:i:3 XO:i:1 XG:i:1 MD:Z:3A3^T2T5C3

[lh3]

Try the latest version of bwa.

[zlu]

Try the latest version of bwa.

Heng,

This was done with BWA 0.5.4.

For resequencing project, does it really matter if the mates are not properly mapped? Can I instead just filter out those reads with low mapping qualities?

Thank you.

[nilshomer]

Perhaps I have misunderstood it but isn't right that properly mapped flag (P string) are only used when read pairs are mapped to the same chromosome with correct insert size? I have about 6% of the properly mapped reads with the P string flag that have mate mapped to different chromosome with 0 insert size as shown below. Has anyone seen this before?

EBRI093151:1:90:555:299#0 pPR1 Chr11 107308221 23 36M Chr12 49 0 TATCCTATTCGAAAGTCGCCATGACCGTGGACATGA BCCBCBBACCB?CBA@BBACCBCAAB<6<?BABBBB XT:A:U NM:i:0 SM:i:23 AM:i:23 X0:i:1 X1:i:1 XM:i:0 XO:i:0 XG:i:0 MD:Z:36

EBRI093151:1:90:555:299#0 pPr2 Chr12 49 23 11S7M1D12M6S Chr11 107308221 0 CTACCGCTTGGGTGGTCATGAATGATTAGCACGCCC AB99@B=BBA>ACCBCCBBCBBCBBBBCBCBB@B@A XT:A:M NM:i:4 SM:i:23 AM:i:23 XM:i:3 XO:i:1 XG:i:1 MD:Z:3A3^T2T5C3

I don't see anywhere in the specification how to set the "properly paired" bit. I would guess this is aligner dependent.

[suseq]

varFilter out put

Hi,

I have used samtools to analyse variations using varFilter. So I have imported an alignment file from BWA in sam format, have sorted and run:
1. samtools pileup -vcf ...
2. samtools.pl varFilter...| awk '$6>=20' ...

It did run but I have problems to interpret all the columns. What I think is:
column 1: chromosome
column 2: first base coordinate from the ref.
column 3: ref. base
column 4: consensus base
column 5: ???
column 6: mapping quality
column 7: ???
column 8: read depth
column 9: read base column
column 10: ???

Does somebody know which values are in column 5, 7 and 10? I could not find this information.

[lh3]

http://samtools.sourceforge.net/cns0.shtml

and

http://samtools.sourceforge.net/samtools.shtml

the pileup command.

[zlu]

I'm wondering for a genome assembly project, will duplicate removal (with samtolls rmdup) and flitering out low mapping quality (e.g mapQ <10) improve my assembly? What do people usually do after mapping with e.g bwa for QC purpose?

Another slightly different issue. Does it matter if 2 fatsq files have the exact identical headers (from 2 solexa runs)? How does samtools sort the bam file? Does it take the header IDs into consideration?

Thank you.