Error in finding SNPs using show-snps

[shiyi-pan]

Hi, Syri is a very good tool, and I've used it for a lot of analysis.

But now I ran into a strange problem，here is my script and ERROR：

python3 syri -c NC_GsojaF.filtered.coords -d NC_GsojaF.filtered.delta -r NC_007942.1.fasta -q GsojaF_polished_final_assembly.fasta

ShV - ERROR - Error in finding SNPs using show-snps: ERROR: 'chloroplast' not found in reference file

Both my reference file and query file only have one sequence named "chloroplast".

Here is my fasta file and mummer ouput files:

syri.zip

I hope you can help me solve this problem in your spare time. I look forward to your reply. Thank you very much.

[lrauschning]

Hi,

seems the issue is coming from the subcall to mummer's show-snps program.
Try converting from .delta info .paf format using e.g. paftools, and/or double-check if the nucmer step worked correctly and the sequence names were always 'chloroplast'.
Hope that helps,

Leon

[shiyi-pan]

Hi,Leon. Thank you very much for your reply. The following is the content in the paf file. Both the reference and query sequences contain the 'chloroplast' sequence:

chloroplast 153275 0 142303 + chloroplast 152218 0 142119 141971 142343 0 NM:i:372 cg:Z:6977M1D17494M6I7283M4I19785M1I163M2D7615M1I3M1I2942M1I13400M1I2590M5I2781M1D27711M1D5168M1I16161M1I1002M1I1220M1I391M1I458M1I105M1I653M1I70M1I806M1I12M1I94M2I84M1I42M1I114M1I63M1I80M1I105M2I127M1I63M1I49M1I73M1I163M1I146M1D142M1I14M1I84M1I268M1I8M1I81M1I279M2I76M1I25M1I35M1I43M1I17M1I28M1I89M1I74M1I15M1I16M1I8M1I78M1I13M1I55M2I11M1I50M1I52M1I49M1I36M1I43M1I65M1D22M1I6M1I13M1I2M1I37M1I39M1I11M1I18M1I16M1I21M1I120M1I6M1I6M1I163M1D5M1I71M1I82M1I15M1I34M1I30M1I29M1D2M1I7M1I16M1I44M1I21M2I4M1I7M1I44M1I25M1D45M1I12M1I110M1I75M1D81M1I29M1I28M1I7M1I58M1I28M1I32M1I16M1I7M1I66M1I56M1I65M1I11M1I116M1I69M1I7M2I6M1I15M1D6M1I33M1D72M1D35M1D21M1I9M1I5M1D11M1I23M1I7M1I5M1D8M1D7M1I9M1D13M1D15M1D9M2I23M1D83M1D2M1I49M1I2M1D13M1I39M1I13M2I15M1D13M1I75M1I16M1I6M1I10M1I32M1D52M1I3M1D48M1I27M1D17M1I25M1I60M1I138M1D10M1I10M1I7M1I18M1I37M1D68M2I14M1D20M1I34M1D67M1D71M1I12M1D104M1I8M1D53M1I98M1D16M1I25M1D15M1D39M1I19M1I32M1D4M1I26M1I5M1I7M1I5M1I11M1I5M1I35M2I15M2I5M2I7M1I21M1I5M1I4M2I17M1I36M1I48M1I37M2I5M1I5M1I12M1I4M1I28M1I9M1I45M3I16M1I4M1I13M2I17M1I3M1I2M1D6M1I7M1I21M2I15M1I3M1I23M4I11M4I6M1I6M2I22M1I13M2I40M4I8M1I14M
chloroplast 153275 144770 153275 + chloroplast 152218 143741 152218 8471 8509 0 NM:i:38 cg:Z:8M2D11M3I8M1D4M1I3M2I5M3I16M1I5M1D30M1I4M1I226M1I989M2I114M1I26M1I40M1I19M1I245M2I183M1I170M1I51M1I21M1I9M1I59M1I285M1I361M1I355M1I764M2I4462M

[lrauschning]

Hi,
sorry, should have been more clear – I meant trying out converrting the .delta to PAF and then running syri with this as input, and seeing if the error persists.

[shiyi-pan]

Thank you for your reply. sorry for reply you late. Here is my script to run syri with paf file.
python3 $PATH_TO_SYRI -c NC_GsojaF.filtered.paf -r NC_007942.1.fasta -q GsojaF_polished_final_assembly.fasta -k -F P --prefix output

The error longer exists, but a new one has occurred.
I set the file type to paf ( -F P ), but syri prompts me to read the bam format and the CIGAR error.

Reading BAM/SAM file - ERROR - Incorrect CIGAR string found. CIGAR string can only have I/D/H/S/X/=. CIGAR STRING: 6977M1D17494M6I7283M4I19785M1I163M2D7615M1I3M1I2942M1I13400M1I2590M5I2781M1D27711M1D5168M1I16161M1I1002M1I1220M1I391M1I458M1I105M1I653M1I70M1I806M1I12M1I94M2I84M1I42M1I114M1I63M1I80M1I105M2I127M1I63M1I49M1I73M1I163M1I146M1D142M1I14M1I84M1I268M1I8M1I81M1I279M2I76M1I25M1I35M1I43M1I17M1I28M1I89M1I74M1I15M1I16M1I8M1I78M1I13M1I55M2I11M1I50M1I52M1I49M1I36M1I43M1I65M1D22M1I6M1I13M1I2M1I37M1I39M1I11M1I18M1I16M1I21M1I120M1I6M1I6M1I163M1D5M1I71M1I82M1I15M1I34M1I30M1I29M1D2M1I7M1I16M1I44M1I21M2I4M1I7M1I44M1I25M1D45M1I12M1I110M1I75M1D81M1I29M1I28M1I7M1I58M1I28M1I32M1I16M1I7M1I66M1I56M1I65M1I11M1I116M1I69M1I7M2I6M1I15M1D6M1I33M1D72M1D35M1D21M1I9M1I5M1D11M1I23M1I7M1I5M1D8M1D7M1I9M1D13M1D15M1D9M2I23M1D83M1D2M1I49M1I2M1D13M1I39M1I13M2I15M1D13M1I75M1I16M1I6M1I10M1I32M1D52M1I3M1D48M1I27M1D17M1I25M1I60M1I138M1D10M1I10M1I7M1I18M1I37M1D68M2I14M1D20M1I34M1D67M1D71M1I12M1D104M1I8M1D53M1I98M1D16M1I25M1D15M1D39M1I19M1I32M1D4M1I26M1I5M1I7M1I5M1I11M1I5M1I35M2I15M2I5M2I7M1I21M1I5M1I4M2I17M1I36M1I48M1I37M2I5M1I5M1I12M1I4M1I28M1I9M1I45M3I16M1I4M1I13M2I17M1I3M1I2M1D6M1I7M1I21M2I15M1I3M1I23M4I11M4I6M1I6M2I22M1I13M2I40M4I8M1I14M. If using minimap2 for alignment, then use the --eqx parameter.

Thank you again.

[lrauschning]

The issue here is that syri expects CIGAR strings reporting exact matches/mismatches (=/X) instead of only aligned (M) tags.
Maybe your delta file does not contain exact match information, or pathtools drops it during conversion – though I think the default is to keep it?
If you don't care about SNPs, you can also just replace every M with a = in the file, though that may cause HDR regions to be annotated as syntenic.

[mnshgl0110]

Hi @shiyi-pan ,

Can you please check whether show-snps is working correctly? Something like this should work:

cat [filename].coords | show-snps -HrTS [filename].delta

Also, syri should be producing multiple *.out files. Are they all empty or contains these two alignments?

[shiyi-pan]

Hi @shiyi-pan ,

Can you please check whether show-snps is working correctly? Something like this should work:

cat [filename].coords | show-snps -HrTS [filename].delta

Also, syri should be producing multiple *.out files. Are they all empty or contains these two alignments?

Here is my script and error:
cat NC_GsojaF.filtered.coords | show-snps -HrTS NC_GsojaF.filtered.delta

ERROR: 'chloroplast' not found in reference file

[shiyi-pan]

Hi @shiyi-pan ,

Can you please check whether show-snps is working correctly? Something like this should work:

cat [filename].coords | show-snps -HrTS [filename].delta

Also, syri should be producing multiple *.out files. Are they all empty or contains these two alignments?

The issue here is that syri expects CIGAR strings reporting exact matches/mismatches (=/X) instead of only aligned (M) tags. Maybe your delta file does not contain exact match information, or pathtools drops it during conversion – though I think the default is to keep it? If you don't care about SNPs, you can also just replace every M with a = in the file, though that may cause HDR regions to be annotated as syntenic.

Thank you for your reply. Unfortunately, this time what I mainly want to obtain is the information of SNPs and indels. The compressed package I uploaded contains the original fasta file. Could you please check where the problem lies or if syri is not suitable for this case. Thank you again.

[shiyi-pan]

Thank you for your reply. Unfortunately, this time what I mainly want to obtain is the information of SNPs and indels. The compressed package I uploaded contains the original fasta file. Could you please check where the problem lies or if syri is not suitable for this case. Thank you again.

[lrauschning]

SyRI can call small variants in well-assembled genomes, including chloroplast genomes.
It seems like something about your .delta file is corrupt. I suggest double-checking your mummer call and/or trying to run the pipeline with a different aligner, e.g. minimap2.
If you only care about small variants and don't expect your assemblies to have structural variation, you can also use bcftools for variant calling from a SAM/BAM alignment.

[shiyi-pan]

Thank you very much , I will try bcftools.