Wiki

How can I report a bug [propose a feature, discuss some aspect of the algorithm]?

Either write a message to the Google group or create an issue in the public Issue Tracker

Is it possible to perform filtering without running the whole pipeline once again?

Yes, one can use parameter --skip-finished to perform filtering for many parameters (like for example number of supporting fragments, split position deviation, homogeneity, etc) once the pipeline was run. The requirement is to have file fusionClusters.annotated.txt in the output folder.

Building InFusion from source code failed. What can be done?

There are multiple factors that influence InFusion code including Linux type, GCC version, available libraries etc. Several issues were created and resolved in fixing this problem.

Here are possible solutions:

If none of them working, create a novel issue.

How to export fusion transcripts from results for PCR primer design?

The easiest way to export the transcript is to use a built-in tool called extract_fusion_sequences from binary folder.

Using this tool it is also possible to extract sequences of fusion transcript having read support. Additionally the tool can extract associated genomic region of certain size and sequences of exons adjacent to fusion breakpoint (transcript recounstruction).

First step to extract the sequences is the isolation of intersting fusions. This can be done using the extractCluster command.

The easiest use case is the fusion sequences derived from SPLIT and BRIDGE read support. This can be done using the following command:

~/infusion/extract_fusion_sequences -ft -d -i selected_fusions.txt -o selected_fusion_transcripts.fa

Extraction of genomic regions adjacent to fusion breakpoint is performed with the same command however one has to provide size of the region and path to the genome sequence:

 ~/infusion/extract_fusion_sequences -ft -l 100 -r /data/scratch/infusion.ens68_data/Homo_sapiens.GRCh37.68.fa  -i vcap_round4.txt  -o vcap_round4_extended.fa

First one has to prepare the fasta file containing the transcripts of genes harbouring the fusion.

#!bash
python ~/playgrnd/infusion/simulation/extractTranscriptsFromGtf.py  --attr-id transcript_name --filter ELK4,SLC45A3 -f /data/genomes/ensemble68/Homo_sapiens.GRCh37.68.fa -g /data/annotations/Homo_sapiens.GRCh37.68.clean.gtf -o transcripts.fa

Next run the tool using the corresponding transcripts:

 ~/playgrnd/infusion/build/Release/extract_fusion_sequences -i elk4_slc45a_iso2.txt -ft -o elk4_slc45a_iso2.fa -gtf /data/annotations/Homo_sapiens.GRCh37.68.gtf -t transcripts.fa