A method using exon RNA-seq counts of
a pair of gene from multiple experiments
to infer if there are fusions between them.

Here present "What is gene fusion":

The reasons for fusion to occur are various:

(wiki) A fusion gene is a hybrid gene formed from two previously independent genes. It can occur as a result of translocation, interstitial deletion, or chromosomal inversion. Fusion genes have been found to be prevalent in all main types of human neoplasia. The identification of these fusion genes play a prominent role in being a diagnostic and prognostic marker.

Briefly,gene fusion is two broken genes join to form a gene and transcribe correspond RNA.Thus the junction of exons from two genes is detectable in RNA sequence.

The importance of gene fusion prediction:

Fusion genes is an important cause of cancer,and therefore a biomarker for cancer diagnosis. Rapid and accurate identification and discovery of fusion genes can not only help to study their role in cancer lesions, but also have an important impact on clinical judgment and treatment of cancer, as many fusion genes can be used as therapeutic targets of targeted drugs.

Here present existing common methods for predicting gene fusion:

With the development of transcriptome sequencing (RNA-seq) technology, gene fusion detection using transcriptome sequencing data has been very common,with various methods and computational tools developed. Those methods align the transcripts to reference genome, then screen out fusion genes by finding spaning reads which have parts located in different genes and split reads covering the junction of the two genes. In the end, filter out the false-positive prediction with strict standards.

Training process: