error:('error: expected one of', {'duplication'}, 'but found', 'deletion'

[Nitin123-4]

I am running mavis and getting the following error:
[2023-04-22 00:02:08] pairing
submitted MP_batch-iF48bHs9AZUjtaxYuuAiS2
MP_batch-iF48bHs9AZUjtaxYuuAiS2 (k9BGiGaU38uASMJ3AeJJor) is SUBMITTED
error:('error: expected one of', {'duplication'}, 'but found', 'deletion', "BPP(Breakpoint(18:43906752R), Breakpoint(18:43906772-43906779L), opposing=False, seq='')", {'tracking_id': 'manta-MantaDEL:196338:0:0:0:0:0', 'library': 'Acute-eosinophilic-leukemia', 'validation_id': 'asQkU4z5QgYCdVyx8wBsGi', 'annotation_id': 'asQkU4z5QgYCdVyx8wBsGi-a1', 'event_type': 'deletion', 'gene1': None, 'gene1_direction': None, 'gene2': 'ENSG00000141622', 'gene2_direction': '5', 'gene1_aliases': None, 'gene2_aliases': 'RNF165', 'gene_product_type': None, 'transcript1': '18:43906752_43906752+', 'transcript2': 'ENST00000593230', 'fusion_splicing_pattern': None, 'fusion_cdna_coding_start': None, 'fusion_cdna_coding_end': None, 'fusion_mapped_domains': None, 'fusion_sequence_fasta_id': None, 'fusion_sequence_fasta_file':
[2023-04-22 00:02:18] summary
Stopping submission. Dependencies not complete
MS_batch-iF48bHs9AZUjtaxYuuAiS2 is NOT SUBMITTED

[Nitin123-4]

Hi team, Please have a look at this request.

[calchoo]

Hi @Nitin123-4, sorry for the delayed response. Would you be able to provide the manta output for that event?

[Nitin123-4]

Hi Calchoo,

Please find attached vcf file.

I think it was:
chr18 43906684 MantaDEL:196338:0:0:0
diploidSV.PASS.vcf.zip
:0:0 TGAGGGCTTCAGCGGCAGAGGGGAGACCGTCCAAGGAGGGCTTCAGCGGCAGAGGGGAGACCGTCCAAG T 261 PASS END=43906752;SVTYPE=DEL;SVLEN=-68;CIGAR=1M68D;CIPOS=0,95;HOMLEN=95;HOMSEQ=GAGGGCTTCAGCGGCAGAGGGGAGACCGTCCAAGGAGGGCTTCAGCGGCAGAGGGGAGACCGTCCAAGGAGGGCTTCAGCGGCAGAGGGGAGACC GT:FT:GQ:PL:PR:SR 1/1:PASS:18:313,21,0:0,0:0,8

ERROR:
error:('error: expected one of', {'duplication'}, 'but found', 'deletion', "BPP(Breakpoint(18:43906752R), Breakpoint(18:43906772-43906779L), opposing=False, seq='')", {'tracking_id': 'manta-MantaDEL:196338:0:0:0:0:0', 'library': 'Leukemia', 'validation_id': 'gZEDaBrudid4n3ZiaWVWzR', 'annotation_id': 'gZEDaBrudid4n3ZiaWVWzR-a1', 'event_type': 'deletion', 'gene1': None, 'gene1_direction': None, 'gene2': 'ENSG00000141622', 'gene2_direction': '5', 'gene1_aliases': None, 'gene2_aliases': 'RNF165', 'gene_product_type': None, 'transcript1': '18:43906752_43906752+', 'transcript2': 'ENST00000593230', 'fusion_splicing_pattern': None, 'fusion_cdna_coding_start': None, 'fusion_cdna_coding_end': None, 'fusion_mapped_domains': None, 'fusion_sequence_fasta_id': None, 'fusion_sequence_fasta_file':

[zhemingfan]

Hi @calchoo and @Nitin123-4,

My 2 cents: Your breakpoints clearly indicate a deletion call, but an interesting observation about the CIPOS position: you have bp1: 43906684(0 CIPOS) and bp2: 43906752 (95 CIPOS) (if I'm reading that correctly). The logic of how MAVIS classifies events is based on start and end coordinates alongside their respective orientations, see: https://mavis.readthedocs.io/en/latest/background/theory/. As the start position of bp2 after accounting for the CIPOS occurs before bp1 (43906657 < 43906684), then it would be a signal expected of a duplication event.

[calchoo]

@Nitin123-4 Are you able to pull that event from the annotation output file? It looks like the positions may have been adjusted in the validate stage as I'm not getting the same breakpoints when running the convert directly on the manta vcf event you posted.

Also what version of mavis are you using?

[Nitin123-4]

Its : mavis version 2.2.11

[calchoo]

Okay as @zhemingfan pointed out it looks like v2.2.11 is converting the event to
chr18 43906684 43906779 L ? None chr18 43906752 43906752 R
while v3.1.1 is converting the event to
chr18 43906684 43906752 L ? None chr18 43906752 43906752 R

Notice the break1 end doesn't go over the break2 end and therefore we shouldn't see this error in the future.
This was addressed in this commit: 7051fc9

@Nitin123-4 are you able to use the latest version (v3.1.1) instead of v2.2.11?

[Nitin123-4]

Thanks for your response. I can see mavis V3 uses snakemake.

Is it possible to fix it in v2.2.11 ? maybe with updated breakpoint.py and vcf.py (7051fc9)

[calchoo]

I could try but it might be faster for you to use the convert from v3 and use the converted file in v2

[calchoo]

@Nitin123-4 This should be fixed in v2.2.12