Nitial set of 83 candidates were selected based on the number ofNitial set of 83

Nitial set of 83 candidates were selected based on the number of
Nitial set of 83 candidates were selected based on the number of paired-end and junction spanning reads as well as each gene taking part in only a few fusions per sample. The final 28 fusion gene candidates were prioritized for laboratory validation based primarily on the number and position of unique short PubMed ID:https://www.ncbi.nlm.nih.gov/pubmed/28914615 read alignment start STI-571 biological activity positions across the fusion junction (Figure 1) and secondarily on location at a copy number transition. One million oligo Agilent aCGH data were combined with sequencing data by drawing images of sequencing coverage and copy number data along with the structure of each candidate gene. Parsing of alignments and other custom analyses were done with in-house developed Python tools. Fusion gene prioritization was done using custom tools built using R [36] and Bioconductor [37].Fusion gene characterizationthat contain the fused exons. A fusion transcript is predicted to be in-frame if any of the transcript-transcript fusions, or their potential splice variants, retain the same frame across the fusion junction. Expression of fusion genes and wild-type parts of the fused genes was calculated as uniquely mapped reads per kilobase of gene sequence per million mapped reads (RPKM). Fusion gene expression was calculated from the number of short reads aligning to the fusion junction. To determine if any of the fused genes has previously been reported to take part in translocations, all 5′ and 3′ genes were compared against the Mitelman Database of Chromosome Aberrations [38]. To determine if fused genes have otherwise been mutated in cancer, all 5′ and 3′ genes were compared against the COSMIC database version 45 [39] and the Cancer gene census [40]. Coverage for each of the fused genes was determined by calculating how many times each nucleotide of the gene was sequenced. Coverage plots were drawn using R [36] and the GenomeGraphs [41] package in Bioconductor [37]. Plots illustrating the discovered fusions and their association to copy number changes were drawn using the Circos software [42].aCGHaCGH was performed as described previously [43] following the protocol provided by Agilent Technologies (version 6), including minor modifications. Briefly, genomic DNA was extracted using TRIzol (Invitrogen) and purified by chloroform extraction and subsequent ethanol precipitation. Three micrograms of digested sample or reference DNA (female genomic DNA; Promega, Madison, WI, USA) was labeled with Cy5-dUTP and Cy3-dUTP, respectively, using Genomic DNA Enzymatic Labeling Kit and hybridized onto SurePrint G3 Human 1M oligo CGH Microarrays (Agilent). To process the data a laser confocal scanner and Feature Extraction software (Agilent) were used according to the manufacturer’s instructions. Data were analyzed with DNA Analytics software, version 4 (Agilent). Raw aCGH data have been deposited in Gene Expression Omnibus [GEO: GSE23949].RT-PCR and quantitative RT-PCRFusion gene frame was predicted by creating all possible fusions between those Ensembl transcripts of both genesThe predicted fusion genes were validated by RT-PCR followed by Sanger sequencing. Fusion junction sequences are listed in Additional file 8. For the RTPCR reactions 3 g of total RNA was converted to firststranded cDNA with random hexamer primers using the High-Capacity cDNA Reverse Transcription kit (Applied Biosystems) according to the manufacturer’s instructions. RT-PCR products were gel-purified (GE Healthcare, Little Chalfont, UK) and cloned into pCRIITOPO clo.