SVDSS call no output

[jiadong324]

Dear author,

Firstly, I got 22 sfs files after running SVDSS search --index ./hg38.fmd --workdir ./HG00733/ --bam HG00733/smoothed.selective.bam. Each of the sfs`` file is named assolutionbatch*.sfs``` and contains some information for each reads.

Then, I run SVDSS call --min-sv-length 50 --workdir ./HG00733 --bam ./HG00733/smoothed.selective.bam --reference ./hg38.no_alt.fa. I got the following output:

SVDSS, Structural Variant Discovery from Sample-specific Strings.
Mode: call
[I] ./HG00733 
[I] PingPong SV Caller running on 4 threads.. 
[I] Loading reference genome from /net/eichler/vol28/eee_shared/assemblies/hg38/no_alt/hg38.no_alt.fa.. 
[I] Extracted chr1 with 248956422 bases. 
[I] Extracted chr10 with 133797422 bases. 
[I] Extracted chr11 with 135086622 bases. 
[I] Extracted chr11_KI270721v1_random with 100316 bases. 
[I] Extracted chr12 with 133275309 bases. 
[I] Extracted chr13 with 114364328 bases. 
[I] Extracted chr14 with 107043718 bases. 
[I] Extracted chr14_GL000009v2_random with 201709 bases. 
[I] Extracted chr14_GL000225v1_random with 211173 bases. 
[I] Extracted chr14_KI270722v1_random with 194050 bases. 
[I] Extracted chr14_GL000194v1_random with 191469 bases. 
[I] Extracted chr14_KI270723v1_random with 38115 bases. 
[I] Extracted chr14_KI270724v1_random with 39555 bases. 
[I] Extracted chr14_KI270725v1_random with 172810 bases. 

There are a number of similar information as shown above. The final output I have is listed below:

[I] Loaded all chromosomes. 
[I] Loading assmbled SFS.. 
[I] Loaded 0 SFS strings on 0 reads. 
[I] Extending superstrings on 4 threads.. 
terminate called recursively
terminate called after throwing an instance of 'terminate called recursively
Aborted

Looking forward to your reply!

Thanks!

[ldenti]

Hi, which version are you using?

I think that you have to rerun the search step using the --assemble argument. We always run the caller on "assembled" SFSs and the caller looks for solution_batch_*.assembled.sfs files in the working directory and not simply solution_batch_*.sfs. Then I also suggest to add --batches 22 to the call command to set how many files to load (this is done automatically in the latest version but not on v1.0.5).

Let me know if this solve the issue. Unfortunately, I still haven't updated the conda recipe to the latest v2 but I updated the README on the repo.. Sorry if this created some confusion on how to run the tool.

Best,

[jiadong324]

Hi,

I am using v1.0.5. Based on your suggestion, I think that I get the correct output, which is svs_poa.vcf.

Thanks!

[jiadong324]

Looking forward to v2.0, it's important to set --batches automatically. BTW, please also update your snakemake file.

Thanks!

[ldenti]

The snakemake should be already updated (do you refer to this one?). Did you encounter any issue?

I just released a new alpha version for SVDSS2 and running some additional tests.

[jiadong324]

Yes, for example

{SVDSS_BIN} smooth --reference {input.fa} --bam {input.bam} --threads {threads} > {output.bam}

It dose not write into standard output, but to your working directory. This is what I used:

SVDSS smooth --reference {ref} --bam {input.bam} --threads {threads} --workdir $( dirname {output.bam} )

Moreover, is that possible to remove the temp results like poa.sam, clusters.txt and etc. They take quite large space if I run for hundreds of samples.

[ldenti]

This snakefile (the one I linked in my previous message) is for SVDSS2. SVDSS2 prints to stdout. And it shouldn't create the temp files by default.

SVDSS1 writes files to the working directory and it creates the temp files. The snakemake for v1 is this one.

Can you please double check that you are using the correct version of SVDSS with the correct Snakefile? Unfortunately v1 and v2 are incompatible.

I'd suggest to use v2 (last alpha) if you don't want the temp files to be stored.