Overall, 130 million pass filter reads were generated for the fresh frozen sample, 190 million pass filter reads for the FFPE sample, and 192 million pass filter reads for the cell line sample. Data was aligned to hg18 assembly of human genome using BWA sequence alignment software (version 0.5.9) and raw alignment BAM files were further processed for quality recalibration, duplicate removal and www.selleckchem.com/products/lapatinib.html local realignment using a custom in-house pipeline based on Picard and GATK tools [18]�C[21]. The alignment statistics are summarized in Tables S1 and S2. For each sample, variants were called from BAM files using samtools and varscan using a minimum coverage cut-off of 10, and only those variants that were called by both algorithms were retained [22], [23].

Fluorescence in-situ Hybridization Fluorescence in-situ hybridizations (FISH) were performed as previously described [14]. Hybridization and post-hybridization washes were done according to the ��LSI procedure�� (Vysis). Hybridizations with the 9p21 (ZytoLight SPEC p16/CEN9 Dual probe, Zytovision) and the Cyclin D1 (ZytoLight SPEC CCND1/CEN11 Dual probe, Zytovision) FISH probes were performed overnight in a humidified chamber at 37��C. All FISH analyses were independently evaluated by two people. Images were obtained by use of an Axioskop 40 fluorescence microscope (Zeiss) equipped with a 63�� objective and an Axiocam MRm camera (Zeiss). Results Flow Sorting of Tumor Populations from Archived FFPE Samples DNA content based flow assays can discriminate cell/nuclei populations based on ploidy including diploid, aneuploid, and elevated 4N(G2/M tetraploid) fractions from fresh frozen biopsies of interest [24].

These assays can be combined with tissue and tumor specific markers to sort subpopulations of diploid and aneuploid populations from routinely collected samples [25]�C[27]. Our previous studies have shown that sorted populations provide optimal templates for high resolution detection of somatic aberrations in each cancer genome [14]. For example homozygous deletions can be detected in aCGH experiments using rigorous objective thresholds (log2ratios 90%) of non-tumor cells. To apply these methods for FFPE samples, thick sections (40�C60 ��m) were initially de waxed, rehydrated in sequential ethanol washes, treated with EDTA then processed with a cocktail of collagenases and hyaluronidase to obtain single nuclei suspensions suitable for flow sorting. For each sample the nuclei were stained with 4,6��-diamidino-2-phenylindole, dihydrochloride (DAPI), disaggregated, and then filtered immediately before analyses on an GSK-3 Influx cytometer (Becton-Dickinson, San Jose CA), with ultraviolet excitation and DAPI emission collected at >450 nm.