Protein Domains from fusions

Question

[Ron]

Hi all,

I came across this post on finding protein domains from gene fusions,so used this software "Oncofuse".

Oncofuse: Prediction Of Driver Gene Fusions From Ngs Data

My input is the STAR-fusion output. My command is following which runs,but the output from Oncofuse is an empty file.

java -Xmx1G -jar $onco_dir/Oncofuse.jar $fusionresults_dir/star-fusion.fusion_candidates.final.abridged starfusion AVG $onco_dir/out_onco.txt

Log file is below:

more oncofuse.sh.o515938.1

Dec 22, 2015 12:20:12 PM org.codehaus.groovy.runtime.m12n.MetaInfExtensionModule newModule
WARNING: Module [groovy-all] - Unable to load extension class [org.codehaus.groovy.runtime.NioGroovyMethods]
[Tue Dec 22 12:20:13 EST 2015] Loading RefSeq data, assuming hg19
[Tue Dec 22 12:20:13 EST 2015] Reading input file, assuming STARFUSION format
Ignoring line #fusion_name    JunctionReads    SpanningFrags    Splice_type    LeftGene    LeftBreakpoint    RightGene    RightBreakpoint
Ignoring line NCOA4--RET    57    35    ONLY_REF_SPLICE    NCOA4^ENSG00000138293.15    chr10:51582939:+    RET^ENSG00000165731.13    chr10:43612032:+
Ignoring line DHRS3--CCDC27    28    14    ONLY_REF_SPLICE    DHRS3^ENSG00000162496.4    chr1:12638746:-    CCDC27^ENSG00000162592.4    chr1:3683099:+
Ignoring line RET--NCOA4    9    6    ONLY_REF_SPLICE    RET^ENSG00000165731.13    chr10:43610184:+    NCOA4^ENSG00000138293.15    chr10:51584616:+
Ignoring line RP11-680G10.1--GSE1    4    2    ONLY_REF_SPLICE    RP11-680G10.1^ENSG00000261567.1    chr16:85391249:+    GSE1^ENSG00000131149.13    chr16:85682158:+
Ignoring line RP11-680G10.1--GSE1    1    2    ONLY_REF_SPLICE    RP11-680G10.1^ENSG00000261567.1    chr16:85391249:+    GSE1^ENSG00000131149.13    chr16:85667520:+
[WARNING] No valid fusions in input file!
[Tue Dec 22 12:20:13 EST 2015] Mapping breakpoints to known genes (this is going to filter a lot)
[Tue Dec 22 12:20:13 EST 2015] 0 fusions mapped out of 0. Dropped fusion candidates: 0 - 5'FPG not mapped, 0 - 3'FPG not mapped, 0 - mapped to same gene, 0 - discordant FPG directions
[Tue Dec 22 12:20:13 EST 2015] Getting features for FPG parts
[Tue Dec 22 12:20:13 EST 2015] Loading expression-related data
[Tue Dec 22 12:20:14 EST 2015] Loading domain and protein interaction interface-related data
[Tue Dec 22 12:20:15 EST 2015] Loading gene ontology data (time-consuming)
[Tue Dec 22 12:20:17 EST 2015] Annotating FPG parts
[Tue Dec 22 12:20:17 EST 2015] =Stage #1: raw data
[Tue Dec 22 12:20:17 EST 2015] ==Domain & PII related
[Tue Dec 22 12:20:17 EST 2015] ==Tissue specific
[Tue Dec 22 12:20:17 EST 2015] =Stage #2: some computations
[Tue Dec 22 12:20:17 EST 2015] ==Domain & PII related (time-consuming)
[Tue Dec 22 12:20:17 EST 2015] ==Tissue specific
[Tue Dec 22 12:20:17 EST 2015] Initializing classifier
[Tue Dec 22 12:20:17 EST 2015] Classifying fusions
[Tue Dec 22 12:20:17 EST 2015] Writing output

Let me know if any suggestions regarding using this software.

Thanks,
Ron

Answer 1

[mikhail.shugay]

Hello! You can try to remove the Splice_type column from the input and run it again. Let me know if it helps. Also check if it runs on https://github.com/mikessh/oncofuse/blob/legacy/src/main/resources/example/example_starfusion.txt file to rule.

[Ron]

Hi,

I removed the Splice type column,it does not work.But the example file you mentioned gives results.It has extra columns

• LeftDistFromRefExonSplice,
• RightDistFromRefExonSplice

probably because of that?

[mikhail.shugay]

Yes, it appears that the example file was from older version/format. Changing the columns so that they are the same as in example file should work.

[Ron]

So I would add two columns with 0's.Hope it does not impact the results.

[Ron]

I added the 2 columns to get the output.Can we get any graphical output?