bwa/picard issue

[cedance]

Hi,

I have a pair of paired end fastq files (lets say forward.fq and reverse.fq) from which I clipped for the adapters, trimmed for quality, and removed barcodes and now have the individual (balanced) files. Off these, I took one of the pairs, lets say, "pe1" and "pe2" and converted to a .SAM file using bwa tool (after indexing corresponding genomic reference etc..).

Then I used ViewSam module of Picard to write "Aligned" and "Unaligned" output files separately. After that when I tried to convert the Unaligned (or Aligned for that matter) file back to fastq (paired end) with "SamToFastq" module of picard, it spits out this error:

Exception in thread "main" net.sf.picard.PicardException: Found 185855 unpaired mates
        at net.sf.picard.sam.SamToFastq.doWork(SamToFastq.java:153)
        at net.sf.picard.cmdline.CommandLineProgram.instanceMain(CommandLineProgram.java:156)
        at net.sf.picard.sam.SamToFastq.main(SamToFastq.java:112)

I tried converting the sam file generated from bwa tool back to fastq and that works!! So this means the ISSUE is with the PICARD ViewSam? I tried validating the Unaligned sam file and got the error log file which consisted of entries like the one shown below 185855 times, which I guess is what the number in the first error represents. I understand what it says, maybe. But I don't understand how or why.

ERROR: Read name SOLEXA1_0503:5:25:1600:20006#0, Mate not found for paired read

How do I fix this error or how did this error occur? I also tried "FixMateInformation" (module from Picard) for both Aligned and Unaligned reads separately and also tried to "MarkDuplicates" (module from Picard).

I am using bwa 0.5.9 and Picard 1.4.1 (both latest from their download pages).

I would appreciate any ideas on fixing this.
Thank you.

[n00c]

Have you tried running Picard with VALIDATION_STRINGENCY=LENIENT?

[cedance]

Hi,

I just tried it. Got the same error!

Exception in thread "main" net.sf.picard.PicardException: Found 185855 unpaired mates

[Seq84]

me too i've the same error.. even with VALIDATION_STRINGENCY=SILENT

[pengchy]

Hi

[Thu Oct 13 00:08:39 CST 2011] net.sf.picard.sam.SamToFastq done. Elapsed time: 0.01 minutes.
Runtime.totalMemory()=12255232
Exception in thread "main" net.sf.samtools.SAMFormatException: Error parsing SAM header. @SQ line missing LN tag. Line:
@SQ     SN:S; ; Line number 2234
        at net.sf.samtools.SAMTextHeaderCodec.reportErrorParsingLine(SAMTextHeaderCodec.java:230)
        at net.sf.samtools.SAMTextHeaderCodec.access$100(SAMTextHeaderCodec.java:39)
        at net.sf.samtools.SAMTextHeaderCodec$ParsedHeaderLine.requireTag(SAMTextHeaderCodec.java:306)
        at net.sf.samtools.SAMTextHeaderCodec.parseSQLine(SAMTextHeaderCodec.java:199)
        at net.sf.samtools.SAMTextHeaderCodec.decode(SAMTextHeaderCodec.java:96)
        at net.sf.samtools.SAMTextReader.readHeader(SAMTextReader.java:198)
        at net.sf.samtools.SAMTextReader.<init>(SAMTextReader.java:79)
        at net.sf.samtools.SAMTextReader.<init>(SAMTextReader.java:88)
        at net.sf.samtools.SAMFileReader.init(SAMFileReader.java:518)
        at net.sf.samtools.SAMFileReader.<init>(SAMFileReader.java:142)
        at net.sf.samtools.SAMFileReader.<init>(SAMFileReader.java:112)
        at net.sf.picard.sam.SamToFastq.doWork(SamToFastq.java:121)

I have checked the line 2234, the line indeed has the LN tag, how was the error produced? Thanks.

[nupurgupta]

I have the same error
' Found unmapped mates'
Any solution to this? No fastq file was generated...

[Naarkhoo]

I have the same issue as following,
Any suggestion ?!

INFO 2013-03-03 21:41:42 SamToFastq Processed 40,000,000 records.
Elapsed time: 00:10:58s. Time for last 1,000,000: 14s. Last read position:
chrX:151,532,960
[Sun Mar 03 21:41:46 EST 2013] net.sf.picard.sam.SamToFastq done. Elapsed time:
11.40 minutes.
Runtime.totalMemory()=4834525184
FAQ: http://sourceforge.net/apps/mediawik...itle=Main_Page
Exception in thread "main" net.sf.picard.PicardException: Found 3494637 unpaired
mates
at net.sf.picard.sam.SamToFastq.doWork(SamToFastq.java:185)
at net.sf.picard.cmdline.CommandLineProgram.instanceMain(CommandLineProgram.java:177)
at net.sf.picard.sam.SamToFastq.main(SamToFastq.java:119)
~

[swbarnes2]

So you just filtered using, say, the 4 flag? So you will have a lot of reads where one end is mapped, and the other is not, right? That seems to be what the software is complaining about.

[aggp11]

A friend recently experienced similar issues with Picard. So he found the following tool to work for him:

http://genome.sph.umich.edu/wiki/BamUtil

You'll have to use it's bam2FastQ tool. Might be worth a try.

[Richard Finney]

Do some accounting. Are all reads paired? or just some. If you have all reads paired, then try "samtools fixmate" first.

If that doesn't work, I hacked fixmate to deal with tougher situations ...

-bash-3.00$ cat bamfixunmaps.c

#include <stdlib.h>
#include <string.h>
#include "bam.h"

/*
usage: ./bamfixunmaps  inputsortebyname.bam outputfixed.bam

remember to resort output bam  by genomic location and re-index 

this takes sorted by NAME (samtools sort -n") input,

how to compile on my machine, fix -I to point to your samtools directory and specify location of libbam.a  and bam.h  ...

vi +19 bamfixunmaps.c ; gcc -Wall -O2 bamfixunmaps.c -o bamfixunmaps -I..  ../libbam.a -lz 

Example fixes for this error message:
ERROR: Record 94341, Read name NCI-GA4_1:2:90:1195:129, Mate unmapped flag does not match read unmapped flag of mate
*/


// currently, this function ONLY works if each read has one hit
void bam_mating_core(bamFile in, bamFile out)
{
    bam_header_t *header;
    bam1_t *b[2];
    int curr, has_prev;

    header = bam_header_read(in);
    bam_header_write(out, header);

    b[0] = bam_init1();
    b[1] = bam_init1();
    curr = 0; has_prev = 0;
    while (bam_read1(in, b[curr]) >= 0) 
    {
    	bam1_t *cur = b[curr], *pre = b[1-curr];
    	if ((cur->core.flag&(BAM_FUNMAP)) || (cur->core.tid < 0))
    	{
             cur->core.qual = 0;
             cur->core.tid  = -1;
             cur->core.pos  = -1; // prints as 0 in sam
             cur->core.flag |= BAM_FUNMAP;  // turn on unmapped bit
    	}
    	if (has_prev) {
    		if (strcmp(bam1_qname(cur), bam1_qname(pre)) == 0) { // identical pair name
    			cur->core.mtid = pre->core.tid; cur->core.mpos = pre->core.pos;
    			pre->core.mtid = cur->core.tid; pre->core.mpos = cur->core.pos;
// rpf
    		if (pre->core.flag&BAM_FUNMAP) // pre is not mapped
                {   // set cur's "minfo"
    		   cur->core.flag |= BAM_FMUNMAP;  // turn on mate unmapped bit
                   cur->core.mtid  = -1;
                   cur->core.mpos  = -1;
                }
    		else
                {
    		   cur->core.flag &= ~BAM_FMUNMAP; // turn off mate unmapped bit
                }
    		if (cur->core.flag&BAM_FUNMAP)     // cur is not NOT mapped
                {
    		   pre->core.flag |= BAM_FMUNMAP;  // turn on unmapped bit
                   pre->core.mtid  = -1;
                   pre->core.mpos  = -1;
                }
    		else
                {
    		   pre->core.flag &= ~BAM_FMUNMAP;  // turn off unmapped bit
                }
// rpf 
    			if (pre->core.tid == cur->core.tid && !(cur->core.flag&(BAM_FUNMAP|BAM_FMUNMAP))
    				&& !(pre->core.flag&(BAM_FUNMAP|BAM_FMUNMAP)))
    			{
    				uint32_t cur5, pre5;
    				cur5 = (cur->core.flag&BAM_FREVERSE)? bam_calend(&cur->core, bam1_cigar(cur)) : cur->core.pos;
    				pre5 = (pre->core.flag&BAM_FREVERSE)? bam_calend(&pre->core, bam1_cigar(pre)) : pre->core.pos;
    				cur->core.isize = pre5 - cur5; pre->core.isize = cur5 - pre5;
    			} else cur->core.isize = pre->core.isize = 0;
    			if (pre->core.flag&BAM_FREVERSE) cur->core.flag |= BAM_FMREVERSE;
    			else cur->core.flag &= ~BAM_FMREVERSE;
    			if (cur->core.flag&BAM_FREVERSE) pre->core.flag |= BAM_FMREVERSE;
    			else pre->core.flag &= ~BAM_FMREVERSE;
    			if (cur->core.flag & BAM_FUNMAP) { pre->core.flag |= BAM_FMUNMAP; pre->core.flag &= ~BAM_FPROPER_PAIR; }
    			if (pre->core.flag & BAM_FUNMAP) { cur->core.flag |= BAM_FMUNMAP; cur->core.flag &= ~BAM_FPROPER_PAIR; }
    			bam_write1(out, pre);
    			bam_write1(out, cur);
    			has_prev = 0;
    		} else { // unpaired or singleton
    			pre->core.mtid = -1; pre->core.mpos = -1; pre->core.isize = 0;
    			if (pre->core.flag & BAM_FPAIRED) {
    				pre->core.flag |= BAM_FMUNMAP;
    				pre->core.flag &= ~BAM_FMREVERSE & ~BAM_FPROPER_PAIR;
    			}
    			bam_write1(out, pre);
    		}
    	} else has_prev = 1;
    	curr = 1 - curr;
    }
    if (has_prev) bam_write1(out, b[1-curr]);
    bam_header_destroy(header);
    bam_destroy1(b[0]);
    bam_destroy1(b[1]);

//    fprintf(stderr,"rpf message %ld fixed by set unmap to qual zero \n", count_unmap_qualNZs);
}

int main(int argc, char *argv[])
{
    bamFile in, out;
    if (argc < 3) {
    	fprintf(stderr, "bamfixunmaps <in.nameSrt.bam> <out.nameSrt.bam>\n");
    	return 1;
    }
    in = (strcmp(argv[1], "-") == 0)? bam_dopen(fileno(stdin), "r") : bam_open(argv[1], "r");
    out = (strcmp(argv[2], "-") == 0)? bam_dopen(fileno(stdout), "w") : bam_open(argv[2], "w");
    bam_mating_core(in, out);
    bam_close(in); bam_close(out);
    return 0;
}

[etwatson]

picard - samtools discrepancy

I am also having this issue, and it appears to be a headache for many people.

I am not interested in paired-end information, since I am comparing unmapped reads to a library of de-novo assembled repeats identified from the raw reads (how much of the mapping issue is due to repetitive elements sampled in the reads?)

1) samtools view -f4 file.bam > unmapped.sam =30,161,064 unmapped reads
2) samtools view -f1 unmapped.sam = 29,415,609 paired, unmapped reads
3) samtools view -F3 unmapped.sam = 752,455 unpaired, unmapped reads

29,415,609 + 752,455 = 30,161,064 unmapped reads. Great, all accounted for.

4) picard-tools SamToFastq on #3 gives me 752,455 reads. Great.
5) picard-tools SamToFastq on #2 gives me 25,317,354 reads and the below error:

SAM validation error: ERROR: Found 4,098,255 unpaired mates.

Why on Earth does samtools tell me I have 29,415,609 paired, unmapped reads while Picard tools tells me that 4,098,255 of those reads are actually UNPAIRED?

coffee break.

[etwatson]

Why on Earth does samtools tell me I have 29,415,609 paired, unmapped reads while Picard tools tells me that 4,098,255 of those reads are actually UNPAIRED?

coffee break.

Ok, after my coffee break, I decided to lean on my awk crutch.

This did the job:

samtools -f4 file.srt.bam | awk 'BEGIN{OFS="\n"}{print "@"$1,$10,"+",$11}'> reads.fastq

[nstoler]

This did the job:

samtools -f4 file.srt.bam | awk 'BEGIN{OFS="\n"}{print "@"$1,$10,"+",$11}'> reads.fastq

Just a caution: You can't get the original reads back by just extracting the 10th column of the SAM.

2 issues result in the 10th column sequences not corresponding precisely to the original reads. Hard clipping results in the sequence stored in the SAM file to be truncated, compared to the original FASTQ read. And secondary alignments result in the same read appearing more than once in the SAM file. See this figure from Heng Li's paper for details on clipping and how reads are stored:

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2723002/figure/F1/

[etwatson]

Just a caution: Hard clipping results in the sequence stored in the SAM file to be truncated, compared to the original FASTQ read. And secondary alignments result in the same read appearing more than once in the SAM file.

Now I'm confused. I am using the -f4 flag to extract unmapped reads. Are unmapped reads altered?

[nstoler]

Now I'm confused. I am using the -f4 flag to extract unmapped reads. Are unmapped reads altered?

The SAM format does not store reads; it stores alignments. The original reads can be reconstructed from the alignments, but it's not as simple as just extracting the sequence in column 10.

For a good example why, take a look at this figure, paying attention to "hard clipping":

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2723002/figure/F1/