Genotyping SVs from other callers

[dr-ashu-geno]

Dear Hannes

Thank you so much for developing such a nice program.

I used GraphTyper on Manta and CNVnator outputs (after getting the union of SVs by svimmer) successfully. However, when running on the data from xTea (a program for calling non-reference mobile element insertions), it throws the following warning for every single insertion position:

[constructor.cpp:782] I do not know how to add an insertion at position 856148

The svimmer output made from xTea VCF files was as follows (one line only for your reference):

chr1 856149 . G INS:ME:ALU 0 . SVTYPE=INS:ME:ALU;SVLEN=259;END=856159;TSD=+GGGATTTTCT;TSDLEN=10;SUBTYPE=two_side_tprt_both;TD_SRC=not_transduction;STRAND=-;LCLIP=1;RCLIP=9;LDISC=12;RDISC=10;LPOLYA=0;RPOLYA=9;LRAWCLIP=3;RRAWCLIP=15;AF_CLIP=12;AF_FMAP=11;AF_DISC=56;AF_CONCORDNT=82;LDRC=0;LDNRC=12;RDRC=0;RDNRC=10;LCOV=47.53;RCOV=30.335;LD_AKR_RC=0;LD_AKR_NRC=12;RD_AKR_RC=10;RD_AKR_NRC=0;LC_CLUSTER=29:29;RC_CLUSTER=280:288;LD_CLUSTER=75.0:281.0;RD_CLUSTER=75.0:210.0;NINDEL=0;CLIP_LEN=4:4:7:21:21:27:26:31:54:55:81:85;INS_INV=Not-5prime-inversion;REF_REP=not_in_Alu_copy;GENE_INFO=exon:ENST00000445118.7:5:LINC01128;NUM_MERGED_SVS=1;STDDEV_POS=0,0;AF=0.521739

At first, GraphTyper threw the error message: unknown SV type: INS:ME:ALU So I changed the SV type using my own script from INS:ME:ALU to INS in field 5, and from SVTYPE=INS:ME:ALU to SVTYPE=INS in field 7. However, the above warning happened, and after one hour no output was made by GraphTyper.

Can you tell me what the warning is related to? Maybe I can modify the svimmer output again so that GraphTyper can get the information it needs to work.

Thank you in advance, Best regards,

[YiweiNiu]

I am also interested in this issue. Did you find a solution? @dr-ashu-geno

[dr-ashu-geno]

I am also interested in this issue. Did you find a solution? @dr-ashu-geno

Unfortunately no! Anyways, if you want to do genotyping for xTea output, there is a program called TypeTE which can do genotyping for ALU insertions only; not for LINE1 and SVA.

Best,

[YiweiNiu]

Thank you very much for your reply @dr-ashu-geno.

Since xTea cannot merge calls from multiple samples. Recently I tried a new tool to call MEI: MEGAnE, which showed better performance than MELT in their preprint paper.