[mikhmv]

Pindel to Dindel

I made improved version of script.

I have a several questions to someone who know how dindel work:
1. Can we replace INV on DEL+INS?
2. Can we do the same for RPL and TD?
3. Do we need to change coordinates (say +1 for INS or DELS)?

Improved code is here:

cat XXXX_D.vcf XXXX_INV.vcf XXXX_SI.vcf XXXX_TD.vcf | awk '/DEL/ {pos=$2+1; print $1,"\t",pos,"\t-"substr($4,2)}; /RPL|INV/ {pos=$2+1; print $1,"\t",pos,"\t-"substr($4,2)"\n"$1,"\t",pos,"\t+"substr($5,2)}; /INS/ {pos=$2+1; print $1"\t"pos"\t+"substr($5,2)}' > XXXX-all.pindel.dindel

[mikhmv]

convert from pindel vcf to dindel format

I ended up with following script:

cat XXXX_D.vcf | grep DEL | awk '{print $1,"\t",$2,"\t&&&"$4}'| sed -e 's/&&&./-/'> XXXX_D.dindel.del
cat XXXX_SI.vcf | grep INS | awk '{print $1,"\t",$2,"\t&&&"$5}'| sed -e 's/&&&./+/' > XXXX_SI.dindel.ins
cat XXXX_D.dindel.del XXXX_SI.dindel.ins > XXXX.dindel.pindel

and added this file (by cat XXXX >>) to Dindel output from the first step.
Regards,
and thanks to Heisman for a help

[Boel]

Dear Dave,

In principle, Pindel works also for overlapping paired-end reads, even for single end reads.

You may need to do the following to achieve a better result:
1. extract reads with bam2pindel.pl (http://www.ebi.ac.uk/~kye/pindel/v_0..._PINDEL.tar.gz) as the internal BAM input function of Pindel may break (or need some changes);
2. modify the coordinate in the extracted records by shifting one read length.
Say read length is 105 and you saw:
@readn
AGGTAATA...
+ chr1 10000 60 200 sample1

change the third line to
+ chr1 9900 60 200 sample1

If you saw "-", put 10100 there, for example.

3. run Pindel (source code at https://trac.nbic.nl/pindel/)

Good luck and send me an email if you have any questions.

Kai

I don't understand what the purpose or effect of this manipulation would be, could someone please explain? Also, is this still necessary in the latest release of pindel? I have overlapping paired end reads and pindel seem to be working on them just fine (so far), I am using the configuration file approach.

Thanks,
Boel

[Heisman]

I have some probelems running pidel and dindel.

pindel2vcf was performed, and get some vcf files, sample_D.vcf and sample_SI.vcf etc.
1) convertVCFToDindel was running,,,,
An error occurred!
Traceback (most recent call last):
File "/home/peshon/bin/convertVCFToDindel.py", line 87, in <module>
main(sys.argv[1:])
File "/home/peshon/bin/convertVCFToDindel.py", line 80, in main
convert(inputVCFFile = options.inputFile, outputVariantFile = options.outputFile, parameters = parameters)
File "/home/peshon/bin/convertVCFToDindel.py", line 39, in convert
if float(dat['QUAL'])>=parameters['minQual']:
ValueError: invalid literal for float(): .
[

2) dindel --analysis realignCandidates was runnning,,,,
There are many same error message.
Could not parse all frequencies in line xxxxx in variants file.
Could not parse any variants in line: xxxxx SKIPPING.

I cannot completely understand for dindel manual 5.3, yet.
Would you tell me a best way of dindel running with candidate variants from pindel output?

I had that same error and was able to fix it by commenting the lines out of the script that dealt with the minQual variable. However, I then had the problem that the script would only work with the pindel SI file, not the D file, and then only a subset of the SI lines (the ones with an "N") would actually be converted.

So, what I just did was to take the pindel SI and D files and use the pindel script to convert them to VCF format. Then I took those files and for each variant in the insertion file (including any with "N"), I took the chromosome, base, and inserted sequence of each line. Ex:

chr1    9325703 .       C       CT      .       PASS    END=9325704;HOMLEN=6;HOMSEQ=TTTTTT;SVLEN=1;SVTYPE=INS   GT:DP   0/1:5

becomes

chr1 9325703 +T

Note the spaces, not tabs. I am always deleting the first character in the 5th column.

For the deletion file, I did the same thing but I only took rows that had the "DEL" tag in the INFO field (the other ones are rearrangements of some sort), and I doing the same thing except inputting a "-" instead of a "+". I still delete the first character except this time I keep the fourth column, not the fifth.

Then, after generating this file, realign those positions to the left most coordinate with the dindel script:

dindel-1.01-src/dindel --analysis realignCandidates --varFile [pindel_variant_list] --outputFile [pindel_variants_realigned] --ref [ref_seq.fa]

Lastly, after the first dindel command, I concatenate the previous output to the dindel output:

dindel-1.01-src/dindel --analysis getCIGARindels --bamFile [aligned_file.bam] --outputFile [first_dindel_output] --ref [ref_seq.fa]

cat first_dindel_output.variants.txt pindel_variants_realigned > new_dindel_variants.txt

And then I used that last file for the rest of dindel.

I do not have any validated indel data to see if this actually improves the indel calls. In my targeted sequence I did go from 291 calls with just dindel to 343 calls when including the pindel calls as described above. I did not do any filtering until the very end when I filtered out variants that were not in my targeted regions.

I can't state for certain this is the right way to do this, but as Pindel does not output quality scores of any type I think this may be the best method. I'd love to get more input from others.

[peshon]

I have some probelems running pidel and dindel.

pindel2vcf was performed, and get some vcf files, sample_D.vcf and sample_SI.vcf etc.
1) convertVCFToDindel was running,,,,
An error occurred!
Traceback (most recent call last):
File "/home/peshon/bin/convertVCFToDindel.py", line 87, in <module>
main(sys.argv[1:])
File "/home/peshon/bin/convertVCFToDindel.py", line 80, in main
convert(inputVCFFile = options.inputFile, outputVariantFile = options.outputFile, parameters = parameters)
File "/home/peshon/bin/convertVCFToDindel.py", line 39, in convert
if float(dat['QUAL'])>=parameters['minQual']:
ValueError: invalid literal for float(): .
[

2) dindel --analysis realignCandidates was runnning,,,,
There are many same error message.
Could not parse all frequencies in line xxxxx in variants file.
Could not parse any variants in line: xxxxx SKIPPING.

I cannot completely understand for dindel manual 5.3, yet.
Would you tell me a best way of dindel running with candidate variants from pindel output?

[KaiYe]

Thank you very much. It works fine.
I still have some questions for the output of pindel:
1. the LI output has a different header and I don't quite understand. The attach is one output. I don't know what does number 4 and 10 means? The insertion size is 121484692-121484686=4? And what's the relationship between the two sequences separated by dash lines?
2. I don't quite understand the INV output. It seems that the sequence in lower case is the inversion, is that right? Is it the reversed sequence of the reference? I don't know why the supporting read doesn't seem to match the sequence in lower case.
3. If I want to confirm breakdancer result, is it ok if I put the output of breakdancer after "-b"? I've tried, but found no similar results.

Thank you very much!

1. The left and right breakpoint of LI (large insertion) are displayed (121484692-121484686). The right coordinate may be smaller than left due to microhomology. The two fields separated by dash lines are reads aligned to the left and right breakpoints, respectively.
2. For INV (inversion), the lower case is the inverted sequences. Can you svn the latest version of Pindel to check whether the alignment is fine. Can you send me the Pindel version number and the picture by email to me ([email protected]).
3. with -b BreakDancer.output, Pindel will try to assemble the breakpoints of them. If you use -Q BreakDancer.Asm, the events with split-read support will be stored in the file BreakDancer.Asm. The assemble efficiency is not 100% and we are working on this. Any comments are welcome.

Kai

[songsy]

It is download item 27 at

https://trac.nbic.nl/pindel/downloads

And you will also find the source code in the svn.

Kai

Thank you very much. It works fine.
I still have some questions for the output of pindel:
1. the LI output has a different header and I don't quite understand. The attach is one output. I don't know what does number 4 and 10 means? The insertion size is 121484692-121484686=4? And what's the relationship between the two sequences separated by dash lines?
2. I don't quite understand the INV output. It seems that the sequence in lower case is the inversion, is that right? Is it the reversed sequence of the reference? I don't know why the supporting read doesn't seem to match the sequence in lower case.
3. If I want to confirm breakdancer result, is it ok if I put the output of breakdancer after "-b"? I've tried, but found no similar results.

Thank you very much!

[KaiYe]

Sorry. I don't know which one contains the pindel2vcf converter on the downloads page. Could you tell me? Thank you very much.

It is download item 27 at

https://trac.nbic.nl/pindel/downloads

And you will also find the source code in the svn.

Kai

[songsy]

We do have a pindel2vcf converter and still testing it. We will make it available tomorrow (May 5 2011) at https://trac.nbic.nl/pindel/downloads.

Cheers,

Kai

Sorry. I don't know which one contains the pindel2vcf converter on the downloads page. Could you tell me? Thank you very much.

[DexterDuncan]

A great big thanks!

Thanks so much for letting me know how to fix this!

Dexter Duncan

[ECO]

In the upper left corner of each post there is an image of a document such as a letter. In one of my posts in the upper left corner there is an image of a hand with the thumb down. When I hover over this with the mouse the pop up says thumbs down. When I hover over the document images the popups say default.

If you notice, upon responding to any the author is given a choice of a few post icons (smiley's, etc). The default is a document image, and the next one is a thumbs down (which is easy to select by accident with a errant tab or arrow-press).

[DexterDuncan]

Red Thumbs Down

Hi Kai,

In the upper left corner of each post there is an image of a document such as a letter. In one of my posts in the upper left corner there is an image of a hand with the thumb down. When I hover over this with the mouse the pop up says thumbs down. When I hover over the document images the popups say default.

Thanks,

Dex

[KaiYe]

What does thumbs down mean in the upper left hand corner of a post as opposed to default?

Thanks,

Dex

Not sure what do you mean. Can you describe further?

Kai

[DexterDuncan]

What does thumbs down mean in the upper left hand corner of a post as opposed to default?

Thanks,

Dex

[KaiYe]

Hi Kai,

could you comment on the options for -R <name_and_version_of_reference_genome> in pindel2vcf?

can I use

-R hg19

, the -h help does not specify much.


Best,

Dave

-R is to specify the reference version while -r is for the path and file name of the reference.

kai