Comprehensive evaluation and validation of targeted next-generation sequencing performance in two clinical laboratories
- Authors:
- Published online on: April 25, 2016 https://doi.org/10.3892/ijo.2016.3497
- Pages: 235-242
Abstract
Introduction
The emergence of next-generation sequencing (NGS) has changed the paradigm of genetic and genomic sequencing studies in personalized medicine. Currently, the power of NGS technology has allowed large scale DNA sequencing projects to be completed in record time, releasing huge amounts of invaluable genetic information. The Cancer Genome Atlas (TCGA) is an international consortium that has sequenced the exome from more than 17,200 tumor patient samples including lung (1–4), colon and rectal (5), breast (6), prostate (7), bladder (8), melanoma (9), gastric adenocarcinoma (10), and numerous other cancer types (11–24). NGS has helped to identify actionable somatic mutations in cancer. Since human cancers are mainly caused by genetic alterations of key drivers or main pathway regulator genes, use of NGS is expected to identify new therapeutic targets and diagnostic markers.
More and more molecular diagnostic laboratories are using NGS to screen mutations of key genes for either drug selection as a targeted therapy or for patient follow-up such as disease recurrence monitoring. The two most common NGS library preparation methods are hybridization or capturing-based (25) and amplicon-based methods (26). Both have proved to work efficiently with DNA extracted from formalin-fixed, paraffin-embedded (FFPE) tissue specimens, which account for the vast majority of samples for molecular and histopathology diagnosis. However, several obstacles prevent the routine implementation of NGS technology in the clinical molecular diagnostic setting: the sophisticated sample preparation process, high cost, time-consuming data analyses, as well as the reproducibility and accuracy of interpretation.
Moreover, it is not always feasible to investigate reproducibility and accuracy of sequencing results among different clinical laboratories due to different systems including technician's skill, DNA and RNA extraction, library preparation, library quality and quantity measurement, sequencing and data analyses. Although the same sequencing library and sequencing reagents are used, the results can be different depending on the several factors mentioned above. This can be a huge problem when considering the significance of the sequencing result and its role as treatment selection.
To address those of the potentially important disparities in workflow, we sought to determine the variations and consistencies between two different clinical laboratories when testing a targeted cancer panel comprising 25 cancer driver genes known to be clinically relevant. Clinical samples and commercially available mutation-positive DNA were tested for laboratory-to-laboratory consistency, validation of intra-laboratory reproducibility, and limit of detection (LOD) analysis. We used both clinical samples and mutation positive DNA control with various types of mutations. We focused on consistency and reproducibility of sequencing results in two different laboratories for the laboratory-to-laboratory accuracy test. We also aimed to test batch-to-batch or person-to-person variations by doing 5 independent sequencings in one laboratory. Finally, we tested the limit of mutation detection by serial dilution of a known mutation-positive DNA sample with a wild-type or non-mutant control sample in a limit of detection test.
Materials and methods
Experimental design
The experimental design of our targeted NGS cancer panel validation consisted of three different experiments conducted at two independent Clinical Laboratory Improvement Amendments (CLIA)-certified laboratories, Encore Clinical (Brisbane, CA, USA; referred to as Laboratory A) and Purity Laboratories (Lake Oswego, OR, USA; referred to as Laboratory B). The first experiment was laboratory-to-laboratory accuracy validation to determine the reproducibility and accuracy of the targeted NGS panel (NextDaySeq™ Pan Cancer HotSpot Panel; CureSeq Inc., Brisbane, CA, USA) of 25 genes selected for clinical relevance of various human cancers. The second experiment was intra-laboratory precision validation to determine the consistency of results in multiple independent experiments in one laboratory. To test intra-laboratory reproducibility, five samples were assayed across five independent experiments by two technicians in the Laboratory B. The third experiment was the limit of detection (LOD) test, which measured the lowest level of mutation allele frequency. It was performed by diluting mutated DNA from a sample with a previously known genotype (KRAS mutation) and mutation allele frequency with DNA from a non-mutated (WT) sample on a wide range of dilutions and assayed in Laboratory B.
Formalin-fixed, paraffin-embedded (FFPE) specimens and samples
In the validation of the three tests mentioned above, we used two commercially available FFPE cell lines, with known genotypes and mutation allele frequencies (EGFR and KRAS Gene-Specific Multiplex Reference Standard cell lines; Horizon Discovery, Cambridge, MA, USA; cat no. HD850 and HD301 for EGFR and KRAS, respectively), 48 FFPE or frozen cancer tissues collected under a protocol (#11-06107) approved by the Committee for Human Research at the University of California, San Francisco, and the AcroMetrix Oncology Hotspot Control DNA (Thermo Fisher Scientific, Waltham, MA, USA; cat. no. 969056) to test multiple mutations across different regions targeted by the NGS panel.
FFPE tissue processing and DNA extraction
For each sample one FFPE tissue section, 5–10 μm in thickness with no more than 2.25 cm2 of tissue area, was used as starting material for DNA extraction with the UltraRapid FFPE DNA extraction kit (CureSeq; cat. no. CS-FFPE-50). Extraction was performed in accordance with the manufacturer's instructions. Briefly, the frosted histological slides containing the FFPE tissue sections were deparaffinized in xylene for 10 min at room temperature in a dedicated fume hood, followed by complete air dry for at least 10 min. The deparaffinized FFPE tissue was hydrated with 5 μl of Solution A (CureSeq) and transferred into a low-binding 0.2 ml PCR tube containing 70 μl of Solution A. The tube was then incubated in a regular thermal cycler for 5 min at 99°C. After adding 10 μl of resuspended Solution B (CureSeq), samples were incubated for 5 min at 60°C followed by 5 min at 99°C. Samples were centrifuged at 1,000 × g for 1 min at room temperature, and the supernatant was transferred into a clean tube without disturbing the pellet.
Quality assessment and quantitation of functional DNA concentration by qPCR
The quantity and quality control of the extracted DNA was evaluated using the DNA Q&Q kit (CureSeq; cat. no. Q&Q-DNA-50) and run on a QuantStudio 6 qPCR platform (Thermo Fisher Scientific), following the manufacturer's instructions. The DNA Q&Q is a fluorescent-based qPCR assay that provides quantitation of human genomic DNA by measuring the presence of short DNA fragments (amplicons called DS1; length ≤85 bp). The integrity of the extracted DNA is evaluated by calculating the ratio of DNA concentration of short (DS1) and long amplicons (called DS2; length <190 bp). Briefly, the extracted DNA was diluted 1:4 (vol/vol) in molecular biology-grade water. Each sample DNA was tested in triplicate for both DS1 and DS2 assays. The qPCR thermal profile and reagents volume shared for DS1 and DS2 were as follows: 2 μl of diluted DNA was mixed with 10 μl of 2× PCR Master Mix, 1 μl of 20× DS1 or DS2 assay oligonucleotides and 7 μl of water. The thermal profile of the qPCR reaction started with an incubation for 2 min at 50°C, followed by 2 min at 95°C and 40 qPCR cycles of 15 sec at 95°C plus 1 min at 60°C. The melting curve consists of a single cycle of 15 sec at 95°C, 1 min at 60°C and a final incubation of 15 sec at 95°C. ROX was used as passive reference. The thresholds of increase in the fluorescence intensity (ΔRn) were 1.29 for DS1 and 1.00 for DS2. The analysis of qPCR data was carried out by using the DNA Q&Q Calculator v7.0 provided by the manufacturer.
Targeted NGS library preparation
The targeted cancer NGS panel (NextDay Seq-Pan Cancer HotSpot Panel kit; CureSeq) was used to prepare the libraries for sequencing in the Ion Torrent PGM sequencer. The libraries were prepared in a three-step protocol. First, 10 ng of DNA was loaded in a multiplexed, high-fidelity PCR reaction, targeting hot-spot mutated loci of the human genome from 25 of the most commonly mutated genes across solid and hematological tumor types (ABL1, AKT1, ALK, BRAF, CTNNB1, DDR2, DNMT3A, EGFR, ERBB2, ESR1, FLT3, GNA11, GNAQ, HRAS, IDH1, IDH2, KRAS, MAP2K1, NRAS, PIK3CA, PTEN, RET, SMAD4, SMO and TSC1) and run for 22 cycles. Second, the PCR products were directly ligated to universal adapters and barcodes. Third, the libraries were purified by using a magnetic bead-based protocol and eluted in 30 μl of 1X LTE buffer. The prepared libraries were stored at 4°C for short-term and at −20°C for long-term storage. Each library (1 μl) was loaded into a High Sensitivity DNA chip (Agilent Technologies, Santa Clara, CA, USA; cat. no. 5067-4626) to evaluate the quality and library yield. The yield from each library (pmol/l) was calculated by running a smear analysis of the electropherogram peaks ranging from 245 to 400 bp, using the Bioanalyzer 2100 platform (Agilent Technologies) and software.
Emulsion PCR and Ion torrent sequencing
Following the manufacturer's instructions, the volume of each of the prepared libraries was adjusted to add equimolar amounts of each library into the emulsion PCR for a final total molarity ranging from 400 to 800 pmoles. The emulsion PCRs were carried out using the Ion PGM Template OT2 200 kit (Thermo Fisher Scientific; cat. no. 4480974) and loaded in the Ion OneTouch™ 2 System, after which non-templated Ion Sphere Particles (ISP) beads were removed by magnetic bead purification (included in the Ion PGM Template OT2 200 kit). After ISP bead enrichment, each library was sequenced using the Ion PGM Sequencing 200 kit V2 (Thermo Fisher Scientific; cat. no. 4482006; lot no. sequencing reagents: 010688; sequencing solution: 010689). The enriched sequencing microreactors from each emulsion PCR were loaded in an Ion 316 v2 Chip or an Ion 318 v2 Chip (Thermo Fisher Scientific; cat no. 4484355; lot no. P32800.1, and P32672.1), and sequenced on an Ion Torrent™ Personal Genome Machine® (PGM) platform.
Laboratory-to-laboratory accuracy test
Forty-eight FFPE or fresh-frozen cancer tissues, two cancer FFPE cell lines samples (Horizon), and one commercial DNA containing multiple cancer mutations (AcroMetrix) were used for targeted NGS analysis. All the library preparation and sequencing were done in Laboratories A and B.
Intra-laboratory precision test
Four cancer FFPE samples (9T, 14T, 19T and 29T) and one normal tissue (3N) were used for the targeted NGS analysis in Laboratory B. Five independent NGS analyses were done by two technicians using the above mentioned five FFPE samples.
Limit of detection (LOD) validation
The LOD study was done at Laboratory B serially diluting a KRAS (G12A) mutant FFPE sample (8T) with a KRAS wild-type sample (27T). The DNA from a mutated sample (8T) was serially diluted with WT DNA to create a decreasing mutation allele frequency titration experiment from 70 to 2%.
NGS data analysis
The PGM's default system, the Ion Torrent Suite (ITS) was used for sequencer data management. ITS integrates a cluster of pre-processing algorithms for NGS data, to align the targeted sequencing data against the human reference genome (hg19) and to normalize the quality scores. The output sequencing data for each barcoded library was aligned against the human genome by the Ion Torrent Suite software and compiled into .bam and their indexed counterpart (.bai) files. These files were used to run coverage analysis of targeted regions and variant calls with DanPA software (CureSeq). DanPA software, by default, only calls mutations with allele frequency >2%. Ion Torrent Suite maximizes the quality of the sequencing data output, filters out the data coming from sequencing microreactors containing polyclonal templates and low quality reads or primer dimers. After the initial filtering of data, the quality of a sequencing run can be assessed by evaluating the percentage of bases with ≥Q20 value, percentage of sequencing reads successfully aligned against the reference human genome, mean of sequencing read length and the median sequencing coverage. The Phred score (Q-score) defines the logarithmic chances that a sequenced nucleotide base is erroneously called, so a base with ≥Q20 indicates that the error rate is <1 in 100.
Statistical analysis
The mutation allele frequency variable was analyzed as a continuous variable. We calculated the correlation between mutation allele frequencies from Laboratory A and B, or from different experiments from intra-laboratory precision validation, by linear regression analysis (GraphPad Prism v 6.0 software). The reproducibility of each mutation detected in the intra-laboratory validation was calculated by coefficient of variation (GraphPad Prism v 6.0 software). All statistical tests were considered significant when the alpha values were <0.05.
Results
Laboratory-to-laboratory accuracy test
The sequencing data obtained was of high quality and consistent inter- and intra-experimentally in both laboratories. The percentage of sequenced bases with ≥Q20 value ranged from 88.18 to 91.98% in Laboratory A and from 89.29 to 92.18% in Laboratory B. The average of sequencing reads aligned against the reference human genome was 96.00% for sequencing runs from Laboratory A and 96.75% for runs from Laboratory B. The median sequencing coverage in targeted regions was of 3,053X and 3,664X for sequencing runs from Laboratory A and B, respectively.
Among the 48 clinical samples used in the accuracy test, a total of 87 mutations were detected, including 63 (72.4%) silent or synonymous and 24 (27.6%) non-synonymous variants, respectively. In the Acrometrix control DNA, 83 expected mutations targeted by the NGS library panel were detected, including 7 (8.4%) silent and 76 (91.6%) non-synonymous variants. In the two commercial FFPE cell lines, a total of 12 mutations were detected, including 2 (16.7%) silent and 10 (83.3%) non-synonymous variants. Out of a total of 182 mutations, 181 were detected in both laboratories. The study of consistency of mutation calls between both clinical laboratories showed that 99.5% (181 out 182) of the mutations detected in Laboratory B were also found in Laboratory A. One synonymous mutation (HRAS H27H; mutation allele frequency, 2.4%), was detected in sample number 25 in Laboratory B but not detected in Laboratory A. Because silent mutations detected in tumor tissue can be either germline polymorphisms or somatic mutations, and the germline variants can affect the degree of correlation between mutation allele frequencies found in both laboratories, two separated analyses were performed: one for all the variants and another for non-synonymous mutations. The regression analysis of mutation allele frequencies from mutations found in both laboratories showed very strong correlation when all the variants (n=181; slope=1.0044; R2=0.996; P<0.0001) were analyzed and if only non-synonymous mutations were analyzed (n=110; slope=0.997; R2=0.9814; P<0.0001) (Fig. 1). Examples of detected mutations in the two laboratories are shown in Fig. 2.
The laboratory-to-laboratory accuracy test showed a sensitivity of 100% (95% CI, 92.45–100%), specificity of 99.97% (95% CI, 99.85–100%), positive predictive value of 97.9% (95% CI, 88.93–99.95%) and a negative predictive value of 100% (95% CI, 99.90–100%).
Intra-laboratory precision test
The quality parameters of the intra-laboratory precision test showed high reproducibility (Table I). Regarding the mutational data, DanPA software detected the same mutation in all five samples across the five independent experiments, showing 100% qualitative reproducibility. Quantitatively, the intra-laboratory reproducibility was very high, as the coefficient of variation of the mutation allele frequency for each mutation across the five independent experiments was always <5% (Table II), regardless of whether the variant was silent or non-synonymous. The regression analysis of mutation allele frequencies found in the five different experiments were highly correlated (Fig. 3).
 | Table IReproducibility of quality parameters of the five sequencing runs from the intra-laboratory precision test. |
| Table IIQuantitative and qualitative reproducibility of mutation calls from the intra-laboratory precision test. |
Limit of detection validation
The LOD experiment was run in the Laboratory B. All the different serial dilutions of the KRAS mutant sample in KRAS WT DNA were run in the same experiment. The quality of the sequencing data was comparable to that from the laboratory-to-laboratory and intra-laboratory validations. The percentage of sequenced bases with ≥Q20 value ranged from 89.14 to 90.58%. More than 97% of the sequencing reads were successfully aligned against the reference human genome. The median sequencing coverage in targeted regions was of 6,350X reads.
The KRAS G12A mutation assayed for LOD was successfully detected at 71.1, 60.8, 41.2, 27.6, 13.3, 6.5, 4.1 and 2.0% mutation allele frequencies, but not in the WT DNA sample, at which the findings ensure the reliability of the data (Table III).
Discussion
Our comparison study of a targeted NGS panel shows data quality and accuracy is very high and satisfactory within and between two clinical laboratories. In the present study, we first conducted a laboratory-to-laboratory accuracy test experiment because reproducibility of a mutation or a variant calling is the basic but the most important factor to be considered for molecular genetic testing in a clinical laboratory. We found that Laboratory B detected 182 synonymous or non-synonymous mutations in a total of 51 samples and that Laboratory A detected 181. One variant calling was not concordant. Our check of the raw sequencing data using a data analysis program and an Integrative Genomics Viewer (IGV; https://www.broadinstitute.org/igv/) showed there was no single mutant read in the data from Laboratory A. The discordant mutation has a low frequency of the mutant allele: 2.4%. This suggests challenging samples with a low minor allele frequency (MAF) must be interpreted carefully. We also did a coverage comparison between two laboratories' results. Although correct mutation detection would be the most important factor for doing NGS analysis, we questioned the consistency of the percentage of mutant alleles in both laboratories. Regression analysis showed a very high correlation of MAF from all the common mutations found in both laboratories (n=110; slope=0.997; R2=0.9814; P<0.0001). The accuracy test result indicates that mutation calling is stable and reliable in two laboratories.
The results of the second test, an intra-laboratory precision test in which we analyzed the outcome of two technicians performing five independent experiments in one laboratory, showed that even in the same clinical laboratory, variations among technicians and reagents or kits could be significant. Thus, intra-laboratory precision validation is important for checking variations caused by batch-to-batch or person-to-person effects. In our experiment, all the variants were detected correctly in five tests (100% reproducibility). The coefficient of variation of the mutation allele frequency across the five independent experiments was <5%.
Our third experiment, a limit of detection (LOD) test, examined the stability of mutation calling in different proportion of tumor part or mutant alleles. We serially diluted a KRAS G12A mutant sample with the KRAS wild-type DNA from ~70% to finally 2% mutation allele frequency. We set the cut-off level for mutation detection at 2% because we believe a mutation candidate with a mutation allele frequency less than 2% has questionable clinical relevance and technical reliability. We limited our scope to regular FFPE tumor or biopsy specimens and aimed to provide coverage of a few hundreds to thousands of reads for selected genes or amplicons. Thus, we believe our cut-off level, 2% in mutation allele frequency, is the minimal and reasonable level for mutation detection. We also selected KRAS mutant for the LOD test sample because KRAS is one of the most frequently mutated genes in various human cancers. In the present study, KRAS G12A mutation was detected in all diluted samples, from 70 through 2%. These findings suggest that most of the FFPE cancer samples with a tumor proportion of at least 10–20% would be suitable for targeted NGS analysis, although screening is always encouraged for samples with a tumor proportion >20%.
Although we believe the consistent results across all three tests are enough to evaluate a targeted NGS cancer panel for most of the clinical samples analyzed in clinical laboratories, several issues remain. First, we tested only FFPE samples and set the cut-off level as 2% of mutation allele frequency. Liquid biopsy or circulating tumor DNA (ctDNA) is becoming popular for early detection of cancer (27) or monitoring tumor recurrence (28). For liquid biopsy specimens, the sequencing coverage should be much higher and the cut-off level should be much lower than for regular tumor specimens. It is therefore important to systemically analyze liquid biopsy specimens in clinical laboratories. Second, we tested tumor specimens with a tumor proportion >20% except for the serial dilution experiment of the LOD test. Many samples from clinics have a tumor proportion <10–20% and tend to be rejected by clinical laboratories or analyzed without thorough consideration. It will be important to have a consensus not only for the clinical meaning but also technical reliability for such samples. Third, tumor heterogeneity or tumor clonal evolution should also be considered when NGS results are interpreted. If multiple different tumor sites from one slide are analyzed separately or a low mutation allele frequency is considered with a very deep sequencing (i.e. more than 10,000X), this issue should be discussed for its clinical meaning before providing a result to cancer patients. Last, more technical aspects beyond NGS should also be thoroughly checked in a clinical laboratory such as DNA extraction methods, storage time or condition of FFPE blocks or slides.
In conclusion, we believe that this present study, done in compliance with the guidelines of American College of Medical Genetics, demonstrates the feasibility of clinical implementation of a targeted NGS cancer panel analysis for personalized medicine.
Acknowledgements
The present study was supported by funds from the CureSeq Inc., and the Thoracic Oncology Laboratory at University of California San Francisco (UCSF). DJM and IJK are equity holders and consultants of CureSeq Inc. We thank Pamela Derish in the UCSF Department of Surgery for editorial assistance with the manuscript.
References
|
Cancer Genome Atlas Research Network. Comprehensive molecular profiling of lung adenocarcinoma. Nature. 511:543–550. 2014. View Article : Google Scholar : PubMed/NCBI | |
|
George J, Lim JS, Jang SJ, Cun Y, Ozretić L, Kong G, Leenders F, Lu X, Fernández-Cuesta L, Bosco G, et al: Comprehensive genomic profiles of small cell lung cancer. Nature. 524:47–53. 2015. View Article : Google Scholar : PubMed/NCBI | |
|
Govindan R, Ding L, Griffith M, Subramanian J, Dees ND, Kanchi KL, Maher CA, Fulton R, Fulton L, Wallis J, et al: Genomic landscape of non-small cell lung cancer in smokers and never-smokers. Cell. 150:1121–1134. 2012. View Article : Google Scholar : PubMed/NCBI | |
|
Hammerman PS, Lawrence MS, Voet D, Jing R, Cibulskis K, Sivachenko A, Stojanov P, McKenna A, Lander ES, Gabriel S, et al; Cancer Genome Atlas Research Network. Comprehensive genomic characterization of squamous cell lung cancers. Nature. 489:519–525. 2012. View Article : Google Scholar | |
|
Network TCGA; Cancer Genome Atlas Network. Comprehensive molecular characterization of human colon and rectal cancer. Nature. 487:330–337. 2012. View Article : Google Scholar : PubMed/NCBI | |
|
Cancer Genome Atlas Network. Comprehensive molecular portraits of human breast tumours. Nature. 490:61–70. 2012. View Article : Google Scholar : PubMed/NCBI | |
|
Abstract G and Brief I; Cancer Genome Atlas Research Network. The Molecular Taxonomy of Primary Prostate Cancer. Cell. 163:1011–1025. 2015. View Article : Google Scholar | |
|
Cancer Genome Atlas Research Network. Comprehensive molecular characterization of urothelial bladder carcinoma. Nature. 507:315–322. 2014. View Article : Google Scholar : PubMed/NCBI | |
|
Hodis E, Watson IR, Kryukov GV, Arold ST, Imielinski M, Theurillat JP, Nickerson E, Auclair D, Li L, Place C, et al: A landscape of driver mutations in melanoma. Cell. 150:251–263. 2012. View Article : Google Scholar : PubMed/NCBI | |
|
Bass AJ, Thorsson V, Shmulevich I, Reynolds SM, Miller M, Bernard B, Hinoue T, Laird PW, Curtis C, Shen H, Weisenberger DJ, et al; Cancer Genome Atlas Research Network. Comprehensive molecular characterization of gastric adenocarcinoma. Nature. 513:202–209. 2014. View Article : Google Scholar : | |
|
Andersson AK, Ma J, Wang J, Chen X, Gedman AL, Dang J, Nakitandwe J, Holmfeldt L, Parker M, Easton J, et al; St. Jude Children's Research Hospital-Washington University Pediatric Cancer Genome Project. The landscape of somatic mutations in infant MLL-rearranged acute lymphoblastic leukemias. Nat Genet. 47:330–337. 2015. View Article : Google Scholar : PubMed/NCBI | |
|
Brennan CWW, Verhaak RGWGW, McKenna A, Campos B, Noushmehr H, Salama SR, Zheng S, Chakravarty D, Sanborn JZ, Berman SH, et al: The somatic genomic landscape of glioblastoma. Cell. 155:462–477. 2013. View Article : Google Scholar : PubMed/NCBI | |
|
Cancer Genome Atlas Research Network. Genomic and epigenomic landscapes of adult de novo acute myeloid leukemia. N Engl J Med. 368:2059–2074. 2013. View Article : Google Scholar : PubMed/NCBI | |
|
Cancer T and Atlas G; Cancer Genome Atlas Research Network. Integrated genomic analyses of ovarian carcinoma. Nature. 474:609–615. 2011. View Article : Google Scholar | |
|
Chan-On W, Nairismägi M-L, Ong CK, Lim WK, Dima S, Pairojkul C, Lim KH, McPherson JR, Cutcutache I, Heng HL, et al: Exome sequencing identifies distinct mutational patterns in liver fluke-related and non-infection-related bile duct cancers. Nat Genet. 45:1474–1478. 2013. View Article : Google Scholar : PubMed/NCBI | |
|
Chen K, Yang D, Li X, Sun B, Song F, Cao W, Brat DJ, Gao Z, Li H, Liang H, et al: Mutational landscape of gastric adenocarcinoma in Chinese: Implications for prognosis and therapy. Proc Natl Acad Sci USA. 112:1107–1112. 2015. View Article : Google Scholar : PubMed/NCBI | |
|
Johnson BE, Mazor T, Hong C, Barnes M, Aihara K, McLean CY, Fouse SD, Yamamoto S, Ueda H, Tatsuno K, et al: Mutational analysis reveals the origin and therapy-driven evolution of recurrent glioma. Science. 343:189–193. 2014. View Article : Google Scholar : | |
|
Jones DTW, Jäger N, Kool M, Zichner T, Hutter B, Sultan M, Cho YJ, Pugh TJ, Hovestadt V, Stütz AM, et al: Dissecting the genomic complexity underlying medulloblastoma. Nature. 488:100–105. 2012. View Article : Google Scholar : PubMed/NCBI | |
|
Lawrence MS, Sougnez C, Lichtenstein L, Cibulskis K, Lander E, Gabriel SB, Getz G, Ally A, Balasundaram M, Birol I, et al; Cancer Genome Atlas Network. Comprehensive genomic characterization of head and neck squamous cell carcinomas. Nature. 517:576–582. 2015. View Article : Google Scholar | |
|
Li M, Zhang Z, Li X, Ye J, Wu X, Tan Z, Liu C, Shen B, Wang XA, Wu W, et al: Whole-exome and targeted gene sequencing of gallbladder carcinoma identifies recurrent mutations in the ErbB pathway. Nat Genet. 46:872–876. 2014. View Article : Google Scholar : PubMed/NCBI | |
|
Lin DC, Hao JJ, Nagata Y, Xu L, Shang L, Meng X, Sato Y, Okuno Y, Varela AM, Ding LW, et al: Genomic and molecular characterization of esophageal squamous cell carcinoma. Nat Genet. 46:467–473. 2014. View Article : Google Scholar : PubMed/NCBI | |
|
Lohr JG, Stojanov P, Carter SL, Cruz-Gordillo P, Lawrence MS, Auclair D, Sougnez C, Knoechel B, Gould J, Saksena G, et al; Multiple Myeloma Research Consortium. Widespread genetic heterogeneity in multiple myeloma: Implications for targeted therapy. Cancer Cell. 25:91–101. 2014. View Article : Google Scholar : PubMed/NCBI | |
|
Creighton CJ, Morgan M, Gunaratne PH, Wheeler DA, Gibbs RA, Gordon Robertson A, Chu A, Beroukhim R, Cibulskis K, Signoretti S, et al; Cancer Genome Atlas Research Network. Comprehensive molecular characterization of clear cell renal cell carcinoma. Nature. 499:43–49. 2013. View Article : Google Scholar | |
|
Tirode F, Surdez D, Ma X, Parker M, Le Deley MC, Bahrami A, Zhang Z, Lapouble E, Grossetête-Lalami S, Rusch M, et al; St. Jude Children's Research Hospital-Washington University Pediatric Cancer Genome Project and the International Cancer Genome Consortium. Genomic landscape of Ewing sarcoma defines an aggressive subtype with co-association of STAG2 and TP53 mutations. Cancer Discov. 4:1342–1353. 2014. View Article : Google Scholar : PubMed/NCBI | |
|
Carrick DM, Mehaffey MG, Sachs MC, Altekruse S, Camalier C, Chuaqui R, Cozen W, Das B, Hernandez BY, Lih CJ, et al: Robustness of next generation sequencing on older formalin-fixed paraffin-embedded tissue. PLoS One. 10:e01273532015. View Article : Google Scholar : PubMed/NCBI | |
|
Fujita S, Masago K, Takeshita J, Okuda C, Otsuka K, Hata A, Kaji R, Katakami N and Hirata Y: Validation of an ion torrent sequencing platform for the detection of gene mutations in biopsy specimens from patients with non-small-cell lung cancer. PLoS One. 10:e01302192015. View Article : Google Scholar : PubMed/NCBI | |
|
Newman AM, Bratman SV, To J, Wynne JF, Eclov NC, Modlin LA, Liu CL, Neal JW, Wakelee HA, Merritt RE, et al: An ultrasensitive method for quantitating circulating tumor DNA with broad patient coverage. Nat Med. 20:548–554. 2014. View Article : Google Scholar : PubMed/NCBI | |
|
Dawson SJ, Tsui DWY, Murtaza M, Biggs H, Rueda OM, Chin SF, Dunning MJ, Gale D, Forshew T, Mahler-Araujo B, et al: Analysis of circulating tumor DNA to monitor metastatic breast cancer. N Engl J Med. 368:1199–1209. 2013. View Article : Google Scholar : PubMed/NCBI |
