Frame issues

[komalsrathi]

Hi,

I have Arriba fusion caller results (run with Gencode v27 equivalent to Ensembl 90) and wanted to test the following out-of-frame fusion:

#gene1	gene2	strand1(gene/fusion)	strand2(gene/fusion)	breakpoint1	breakpoint2	site1	site2	type	direction1	direction2	split_reads1	split_reads2	discordant_mates	coverage1	coverage2	confidence	closest_genomic_breakpoint1	closest_genomic_breakpoint2	filters	fusion_transcript	reading_frame	peptide_sequence	read_identifiers	sample_name
CHMP1A	DPEP1	-/-	+/-	16:89651569	16:89625844	splice-site	intron	deletion/read-through/5'-5'	upstream	downstream	7	6	1	118	36	high	.	.	duplicates(4),low_entropy(1)	GCTTGGTCGGTTCGATCGCCGCCGGGACCTGACACCGCCCGGAGTTGGCGTCCCTTCTCCCTCTCCGAGTGCTGCTCCTGTCATTGTGGCCATGGACG___ATACCCTGTTCCAGTTGAAG___TTCACGGCGAAGCAGCTGGAGAAGCTGGCCAAGAAGGCGGAGAAGGACTCCAAGGCGGAGCAGGCCAAAGTGAAGAAG|TCTCTTCGTCTGTCAAGTAGAAGTCATGGATGTACCTACTTTACGACAGTCCTGCTGTGAGCACAAAGGTACCAGCCACTCAGATCACCCTCGGATCAAG___CTCTGTCTCCCAGAACCACACAGAAGCCCCATCCACAGCCAACATGCAGACGCCACCGTGGCCATTTCGGATGATACCACTTCTGCACAGGCAGCCACAGGCAATGCGGACCTACCCACGCCCCCCACACACAGATCATCGCAG	out-of-frame	MDDTLFQLKFTAKQLEKLAKKAEKDSKAEQAKVKK|slrlssrshgctyfttvll*	.	0062396f-103a-4c25-bfdc-a746782768d9

I ran agfusion like this:

agfusion annotate \
--gene5prime CHMP1A \
--gene3prime DPEP1 \
--junction5prime 89651569 \
--junction3prime 89625844 \
-db agfusion.homo_sapiens.90.db \
-o CHMP1A-DPEP1

I got quite the opposite result from agfusion, predicting it to be an in-frame fusion:

$ cat CHMP1A-DPEP1.fusion_transcripts.txt
5'_gene,3'_gene,5'_transcript,3'_transcript,5'_strand,3'_strand,5'_transcript_biotype,3'_transcript_biotype,Fusion_effect,Protein_length,Protein_weight_(kD)
CHMP1A,DPEP1,ENST00000397901,ENST00000393092,-,+,protein_coding,protein_coding,in-frame,446,49.6940567

I am not sure why that is happening, any ideas/suggestions would be much appreciated.