SVDetect - No links have been found with the selected type of structural variation

[tarunmishra.rsg]

Hi all,

I am trying to use SVDetect to predict SVs for a exomic data. I have followed the following command

SVDetect_r0.5/bin/SVDetect linking filtering links2SV -conf test_sample.sv.conf

My test_sample.sv.conf contains requisite parameters.

The tool runs successfully but in the end I get

# Linking procedure...
-- file=/home/tarun/svdetect/test.ab.sam
-- type=all
-- window size=1000, step length=250
-- Program stopped!
-- No links have been found with the selected type of structural variations (all)

My sample insert size is 500, read length 76, illumina Paired end data. Please suggest what could be done. and possible reason for no detection of SVs.

Thanks a lot
regards
Tarun

[Jon_Keats]

If your data is an exon capture dataset you will likely not have many structural variations as most will be in the intergenic and intronic regions you selected against by doing exon capture.

[tarunmishra.rsg]

Hi Jones , thanks for your reply

I tried using the SVdetect now using a Matepair sample (WG) . This time as well i got the same prompt
-- No links have been found with the selected type of structural variations (all).

I am pasting the .conf.file parameters that I have used. Please let me know if there is any error due to the choice of these parameters. I have calculated the mu and sigma values using the max, min insert size and used the values of mu and sigma to calculate the value of window size (2*mu+2*sqrt (2*sigma)) and step length (1/4 of window size)as given in the manual.

command used is
SVDetect linking -conf test_data.sv.conf

conf file is:
<general>
input_format = sam
sv_type = all
mates_orientation=RF
output_dir=Test
</general>

<detection>
tag_length=5000
read1_length=50
read2_length=50
window_size=6331
step_length=1582
mates_file=test.ab.sam
cmap_file=hs18.len
</detection>

My aim to detect CNV, SV, Circos output and comparison for different samples. I also wish to know that should I give these commands (linking, filtering,links2SV etc) individually or I can give all at once in same command line. If I do the former, do i need all the blocks in the conf file. I tried only with linking first, and gave only detection block as you see. Also what does tag_length refers to, Could not find much in the manual, but got as an error during execution(tag_length not defined), so randomly gave a value, that in fact made the program run.

But got the result as above mentioned. Plz suggest.

thanks a lot
tarun

[Jon_Keats]

I'm assuming you did the test runs bundled in the SVDetect download to make sure the application is running correctly in your enviroment.

Couple of questions

Assuming you have Illumina data given the RF format, how did you align, does it do pairing correctly in the RF format?

Tag_length= the read length.
Were did you get the read1_length and read2_length variables as those just duplicate the Tag_length. Moreover, the tag_length of 5000 is likely the source of your problems.

[ritzriya]

Same error, No solution!

Hi,

I am facing the same error as tarunmishra.rsg. Following is my config file:

<general>
input_format = bam
sv_type = all
mates_orientation = RF
output_dir =output
read1_length=35
read2_length=35
mates_file=output/chr13only_normal_sort.ab.bam
cmap_file=hg19_chr13.len
tmp_dir=temporary
</general>

<detection>
split_mate_file=1
window_size=4000
step_length=1500
</detection>

It gives me the following error:

# Splitted mate files of "//home/pratibha/softwares/CNVtools/SVtools/SVDetect_for_SV/SVDetect_r0.8/scripts/output/chr13only_normal_sort.ab.bam" created.
# interchrs : Linking procedure...
-- file=temporary/mates/chr13only_normal_sort.ab.bam.all.interchrs
-- chromosome=interchrs
-- input format=sam
-- type=all
-- read1 length=35, read2 length=35
-- window size=500, step length=20
-- interchrs : No mate-pairs !
-- interchrs : No links have been found with the selected type of structural variations (all)
# 13 : Linking procedure...
-- file=temporary/mates/chr13only_normal_sort.ab.bam.all.13
-- chromosome=13
-- input format=sam
-- type=all
-- read1 length=35, read2 length=35
-- window size=500, step length=20
-- 13 : No mate-pairs !
-- 13 : No links have been found with the selected type of structural variations (all)
ls: cannot access temporary/links/chr13only_normal_sort.ab.bam.all*links: No such file or directory
# Error: No links files created at temporary/ : at ./SVDetect line 155, <MATES> line 3266914.

I am using SVDetect 0.8 version, which is the latest. Any help would be really great to solve this issue.

Thanks in advance.

[shruti]

Hi.. I'm using SVDetect _r0.8 and
I'm also getting the same error
-- interchrs : No mate-pairs !
-- interchrs : No links have been found with the selected type of structural variations (all)
Let me know if you were able to get around this problem.

[zeib]

Dear shruti

About your error, check the content of your chromosome length file (.len) and if the chromosome names in this file correspond to the chromosome names listed in your alignment BAM file.

Bruno

[shruti]

Hi Zeib,
Thanks.. it seems to work now..

[Qingl]

Hi,
I am using SV_detect (Version: 0.7m) to analyze simulated illumina paired end 100 bp reads with 300 bp fragment size - using bwa to align the 16x coverage depth reads back to human chr1(ref). and I am getting similar error,,,
# Making the fragments library...
-- 1
# Linking procedure...
-- file=sample_ab_mates.sam
-- input format=sam
-- type=all
-- read1 length=100, read2 length=100
-- window size=900, step length=225
-- Program stopped!
-- No links have been found with the selected type of structural variations

Does anyone ever make this software work for paired_end reads?
Thanks in advance!
Q

[Qingl]

Hi,
I am using SV_detect (Version: 0.7m) to analyze simulated illumina paired end 100 bp reads with 300 bp fragment size - using bwa to align the 16x coverage depth reads back to human chr1(ref). and I am getting similar error,,,
# Making the fragments library...
-- 1
# Linking procedure...
-- file=sample_ab_mates.sam
-- input format=sam
-- type=all
-- read1 length=100, read2 length=100
-- window size=900, step length=225
-- Program stopped!
-- No links have been found with the selected type of structural variations

Does anyone ever make this software work for paired_end reads?
Thanks in advance!
Q

Hi,
I have to admit that it's the same problem- chromosome names in chromosome length file must exactly correspond to the chromosome names listed in the alignment SAM file, it's resolved for paired end reads also...

[Lien]

Hi all,

I'm trying to use SVDetect. However, when I start 'linking' I get the following error:

Linking procedure...
-- file=sample.ab.bam
-- input format=sam
-- type=all
-- read1 length=100, read2 length=100
-- window size=451, step length=112
[sam_header_read2] 93 sequences loaded.
[sam_read1] reference 'GO:none' is recognized as '*'.
Parse error at line 1: invalid CIGAR character
-- Program stopped!
-- No links have been found with the selected type of structural variations (all)

Apparently there is something wrong with the header of my bam-file. But this bam-file was created using the BAM_preprocessingpairs.pl script included in the SVDetect package. Does anyone know how to solve this problem?

Many thanks,
Lien