Fusion gene detect tools for Solid (colorspace)single end 50bp RNA-seq data

[yinxiaohe]

Does someone know which tool can detect fusion gene form Solid single end 50bp RNA-seq reads. I found several softwares such as FusionMap,FusionSeq and so on. But some need paired end reads and some cant't handle colorspace reads. Now I still didn't find a tool to do the job with my data. Can someone help me. Thank you very much!

[louis7781x]

Although there is no software to detect fusion gene in coloerspace now.

You can do as following,
First,using tophat to run your data .
Second,use function of FusionMap:--pereport to select discordant read pairs from sam file.
Third,use GSNAP to discovery junction site.

Final,combined report from discordant read pairs and junction site.

[jochensupper]

With single-end reads you could in principle try to detect gene fusions from spliced-alignments. There are mappers supporting this in color space, for instance, the Genomatix mapper (FYI I work for Genomatix).

However, I would not recommend this strategy at all. You will get lots of false positive predictions. Especially since your reads are only 50 bps long, so even in an ideal case you will only have 25 bps on each side of the split. This will not guarantee correct mappings with high precision.

When predicting gene fusion candidates we use spliced-alignments to determine exact fusion breakpoints. But we only do this after having identified the gene fusion candidates using paired-end information.

[yinxiaohe]

Maybe single end 50bp data are really not suitable to detect fusion gene, somehow I'll have a try as you two suggested. Thank you very much!