cBioPortal Import Error: Duplicate SV/Fusion Entries

[Setlem, Rohit, M.S.]

Hi Team,

We are attempting to upload RNA fusion SV data to cBioportal and encountering the following error:

ERROR: data_fusions_extended.txt: lines [130, 131, 132, (12417 more)]:

Duplicate entry in structural variant data; values encountered:

[ ... Results (already defined on line 117) … ]

Below are some example entries from the file that trigger the error:

Sample_Id

SV_Status

Site1_Hugo_Symbol

Site1_Ensembl_Transcript_Id

Site1_Entrez_Gene_Id

Site1_Region_Number

Site1_Region

Site1_Chromosome

Site1_Contig

Site1_Position

Site2_Hugo_Symbol

Site2_Ensembl_Transcript_Id

Site2_Entrez_Gene_Id

Site2_Region_Number

Site2_Region

Site2_Chromosome

Site2_Contig

Site2_Position

Site2_Effect_On_Frame

NCBI_Build

Class

RNA-Control-2

SOMATIC

VCL

NM_014000

VCL

16

exon

chr10

NA

75865111

NTRK2

Refseq_3

NTRK2

12

exon

chr9

NA

87356807

In-Frame

GRCh37

FUSION

RNA-Control-2

SOMATIC

VCL

NM_014000

VCL

16

exon

chr10

NA

75865111

NTRK2

Refseq_3

NTRK2

13

exon

chr9

NA

87359888

In-Frame

GRCh37

FUSION

In this case, the fusion is between the same gene pair in the same sample (VCL-NTRK2), but the 3’  breakpoint differs (chr9:87356807 vs chr9:87359888) and the region number differs (12 vs 13).

We also see similar failures in other instances where the gene pairs are same, but the transcript identifiers vary, and the importer reports them as duplicates.

Could you please clarify which fields are used to define duplicates for SV/fusion entries during import? 

Additionally, what is the recommended approach to represent the same gene pair within a sample but with different breakpoints or transcripts ? 

We would appreciate your guidance on how to handle these situations to ensure a successful upload the biologically distinct events

Thanks,

Rohit Setlem | Senior Bioinformatician | Quantitative Health Sciences | Email: setlem...@mayo.edu

Mayo Clinic | 200 First Street SW | Rochester, MN 55905 | www.mayoclinic.org

 

[Benjamin Gross]

Hi Rohit,

Thank you for the email. validateData.py is the code that is generating this error. The validator is called by the meta importer before the actual import. The columns below are taken into account when determining SV record uniqueness (see https://github.com/cBioPortal/cbioportal-core/blob/main/scripts/importer/validateData.py#L3151). If the column headers in the file are not named exactly as shown below, they may be ignored by the validator when checking for uniqueness. I would check that first.

 UNIQUENESS_COLUMNS = [

        'Sample_Id',

        'Site1_Hugo_Symbol',

        'Site1_Entrez_Gene_Id',

        'Site1_Chromosome',

        'Site1_Position',

        'Site1_Region_Number',

        'Site1_Ensembl_Transcript_Id', 

        'Site2_Hugo_Symbol',

        'Site2_Entrez_Gene_Id',

        'Site2_Chromosome',

        'Site2_Position',

        'Site2_Region_Number',

        'Site2_Ensembl_Transcript_Id',

        'Event_Info’]

Let me know how it goes.

-Benjamin

--
You received this message because you are subscribed to the Google Groups "cBioPortal for Cancer Genomics Discussion Group" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cbioportal+...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/cbioportal/SJ2PR01MB848445781B2915BCDAD511D2FA94A%40SJ2PR01MB8484.prod.exchangelabs.com.