Research ArticleExperimental Studies

Robustness of a Cancer Profiling Test Using Formalin-fixed Paraffin Embedded Tumor Specimens

SHIN ITO, IKURO SATO, MAI MOCHIZUKI, KAZUNORI YAMAGUCHI, KEIICHI TAMAI, TAKAMICHI MINATO, NOBUHIRO TANUMA, HIROSHI SHIMA and JUN YASUDA
Anticancer Research March 2021, 41 (3) 1341-1348; DOI: https://doi.org/10.21873/anticanres.14891

Abstract

Background/Aim: Cancer profiling tests using formalin-fixed paraffin-embedded (FFPE) specimens with various conditions have become an essential tool for cancer treatment. The robustness of these tests needs to be addressed. Materials and Methods: A cancer profiling test, NCC oncopanel, was tested with FFPE specimens from various tissues with different storage conditions and fixation lengths. Next generation sequencing was performed with Miseq and the data were assembled using the human reference genome hg19. Results: Duration of storage and fixation affected the mapping statistics. Prolonged storage increased outward read paring and longer fixation rates caused increased singletons and unmapped reads. Conclusion: Our results indicate that a cancer profiling test with target capturing method, NCC oncopanel, shows robustness for FFPE cancer specimens with various storage conditions.

Key Words:

Next generation sequencing (NGS) cancer profiling tests for surgical specimens is an essential tool for cancer treatment. In Japan, the national health insurance system recently approved the use of two cancer profiling tests, FoundationOne CDx (1) and NCC oncopanel (2), at government-certified hospitals. The costs of these cancer profiling tests are high and can only be covered once in the lifetime of a patient by the national health insurance in Japan. Therefore, the robustness of the cancer profiling test for a given specimen quality is critical.

The NCC oncopanel is a hybridization capture based NGS assay designed to examine mutations, amplifications and homozygous deletions of the entire coding regions of 114 genes of clinical or preclinical relevance along with rearrangements in 12 oncogenes. The present version of NCC oncopanel (version 4) is widely used as a cancer profiling test in the Japanese hospitals certified by the Japanese government. Version 2 of the panel is provided for research use and we can test the effect of storage of formalin-fixed paraffin-embedded (FFPE) specimen on the target capturing cancer profiling tests in a research laboratory setting.

Various clinical factors can affect the quality of FFPE specimens, from surgical procedures to the sectioning of paraffin blocks (3). For example, samples fixed in unbuffered formalin for a long period (>7 days) may not be suitable for molecular analysis (4). It is also known that formalin can induce artificial mutations via cytosine to thymine (C to T) mutations and that these artificial mutations can be reduced by uracil-DNA glycosylase treatment (5, 6).

It is important to understand the degree to which the output of cancer profiling tests is affected by the variations in FFPE sample quality in a clinical setting. Therefore, we investigated the mapping quality of the output of a cancer profiling test, NCC oncopanel (experimental version). We found that the NCC oncopanel showed robustness regarding the length of storage and fixation time for representative variants of the tumor tissue.

Materials and Methods

FFPE samples. We selected tissues of interest from our archive at the Miyagi Cancer Center Hospital (Table I). Surgically removed cancer tissue was treated with a 10-fold diluted 37% formaldehyde solution (069-0047; Fujifilm Wako Pure Chemical Co., Osaka, Japan). Only sample 15 was fixed with 10% neutral buffered formalin solution (060-01667; Fujifilm Wako Pure Chemical Co.). Paraffin embedding and preparation of slides were done according to standard protocols. This study was approved by the in-house ethical committee of the Miyagi Cancer Center (registration number 2018-069).

View this table:
Table I.

Summary of samples.

DNA extraction from FFPE specimens. Tissues were cut from paraffin blocks (5 μM thick, 10 pieces) and DNA was extracted with the Maxwell RSC DNA FFPE Kit - PKK, Custom (AX2500; Promega Corporation, Madison, WI, USA) and Maxwell RSC Instrument (Promega Corporation), according to the manufacturer’s instructions.

Cell culture and cellular DNA extraction. PANC-1 human pancreatic carcinoma cells were cultured in RPMI-1640 medium (189-02025; Fujifilm Wako Pure Chemical Co.) containing 9% fetal bovine serum (SFBM30; Equitech-Bio, Kerrville, TX, USA). Cells were incubated at 37°C with 5% CO2 and collected for DNA extraction at 60% confluence in a 10-cm dish. DNA was extracted from cells using a DNeasy Blood & Tissue Kit (69504; Qiagen Gmbh, Hilden, Germany) according to the manufacturer’s instructions.

Measurement of extracted DNA concentration and quality control. A Qubit dsDNA HS Assay Kit (Q32851; Thermo Fisher Scientific, Waltham, MA, USA) and a Qubit Fluorometer (Thermo Fisher Scientific) was used to measure the concentration of DNA extracted from surgical specimens and cultured cells. We also used the Agilent NGS FFPE QC Kit (G9700A; Agilent Technologies, Santa Clara, CA, USA) and LightCycler 480 System (Roche Diagnostics, Indianapolis, IN, USA) for quantitative polymerase chain reaction (qPCR) assessment of DNA concentration and sample DNA fragmentation. qPCR was performed according the manufacturer’s instructions. DNA concentrations of the samples with a ΔΔCq value less than 1 were based on the DNA concentration measured by the Qubit.

Library preparation for NGS. 200 ng of each DNA sample was used for library preparation. DNA was fragmented with SureSelect XT HS and XT Low Input Enzymatic Fragmentation (5191-4079; Agilent Technologies). SureSelect XT HS Reagents (G9702A, Agilent Technologies) and SureSelect NCC oncopanel (931195; Agilent Technologies) was used for the preparation of the NCC oncopanel capture library for the Illumina Next Generation sequencers. The TapeStation 2200 system (Agilent Technologies) and D1000 ScreenTape and its reagent (5067-5582 and 5067-5583, respectively) were used for pre-capture and post-capture library size checks and quantification. All processes from DNA fragmentation to library quality control were performed according to the manufacturer’s instructions.

NGS sequencing. The prepared NGS libraries were diluted to 8 nM each based on the quantitative results from TapeStation. A library pool was created by mixing equal amounts of these libraries. Paired end sequencing was performed with the MiSeq Reagents Kit v2 300 Cycle (MS102-2002; Illumina, Inc., San Diego, CA, USA) and MiSeq (Illumina) in 151-cycles run.

Bioinformatics. The raw reads were trimmed of adapter sequences and low-quality bases by Trimmomatic 0.39 with default options (7). Only paired reads were mapped onto the hg19 reference genome with BWA-MEM (8) and the generated sam files were sorted and compressed with samtools (9). The mapping statistics data were obtained with samtools stat and flagstat options. Variant calling was then performed with Mutect2 in the Genome Analysis Toolkit (GATK), version 4.1.4.1. The filtering conditions were as following: false-discovery-rate=0.01, unique-alt-read-count=20, min-allele-fraction=0.1, minimum depth=100, and within the regions defined by 0471501_Padded.bed in the NCC oncopanel target region [(911 kb)+100 bp flanking sequences] (1.4 Mb total).

The annotations for the called variants were performed with Annovar (10) with a custom-made 4.7 KJPN variant dataset and COSMIC 90 (11). The 4.7KJPN variant data, consisted of whole genome sequencing variant data of more than 4,700 Japanese individuals, were downloaded from the jMorp database (https://jmorp.megabank.tohoku.ac.jp/) (12). Variant extraction was performed with bcftools (9, 13). The variants were also annotated with the COSMIC database version 90 (11).

Verification of variants. Prrimers were designed to amplify the regions containing the 12 mutations and 4 indels selected for validation (primer sequences are available upon request). The targeted regions were amplified by PCR with KOD FX (KFX-101; TOYOBO CO., Osaka, Japan), PCRx Enhancer System (11495017; Thermo Fisher Scientific), and the corresponding primer sets. After denaturation at 94°C for 2 min, 35 cycles of denaturation at 98°C for 10 s, annealing at 55°C for 30 s, and extension at 68°C for 30 s were performed, followed by a 7-min extension reaction at 68°C. After PCR amplification, PCR products were subjected to ExoSAP-IT Express PCR Cleanup Reagents (75001; Thermo Fisher Scientific) and processed at 37°C for 4 min and at 80°C for 1 min to remove single-strand DNA and dNTPs. Sanger sequence analysis of each PCR product was performed by Eurofins Genomics K.K. (Tokyo, Japan). Electrophoresis was performed with NuSieve GTG Agarose (50081; Lonza, Rockland, ME) and Agarose S (316-01191; Nippon Gene, Tokyo, Japan) 3:1 mixture gel for verification of large indels. The presence or absence of indels in the PCR products was verified by the differentiation of electrophoretic mobility. The molecular marker was Quick-Load 100 bp DNA Ladder (N0467; NEB, Ipswich, MA, USA).

Results

The effects of the storage and fixation lengths of FFPE samples on genomic DNA quality may vary. The initial aim of the present study was to identify “thresholds” for the cancer profiling test regarding the storage and fixation periods for FFPE samples. Table I summarizes the conditions of the samples tested in this study. The storage and fixation durations ranged from 1-5 years and 1-14 days, respectively. Quick estimates of the genomic DNA quality of FFPE samples can be undertaken by quantitative PCR. A higher ΔΔCq value indicates more damaged DNA templates. The same storage and fixation times resulted in a variety of ΔΔCq values among the samples of different origins, indicating that a number of clinical factors may affect the quality of genomic DNA extracted from FFPE specimens (Table I). The storage length (years) are more correlated to ΔΔCq than the fixation length (days) for all samples analyzed in this study (Pearson’s correlation coefficients are 0.70 and 0.26, respectively).

Prolonged fixation and storage of FFPE samples may affect the mapping quality of the NGS output. The obtained sequence reads from damaged genomic DNA showed high-quality base calls, while the mapping data of the sequence reads to the reference genome may have been affected by formalin fixation-derived artifacts. Table II summarizes the relationship between the mapping statistics and the condition of the FFPE samples. The coefficient variables (CVs) among samples normalized with the number of total sequencing reads (Table II) were used for comparison with the length of storage (years) or fixation (days) or the ΔΔCq value of the samples, regardless of their origins (Table II). Some of the mapping statistics were very invariable; the CVs of “Reads properly paired”, “Bases mapped (cigar)”, and “Reads mapped and paired” were less than 0.005 (Table II). The CVs of “Singletons”, “Unmapped reads”, “Non-primary alignments”, “Outward oriented pairs”, and “Pairs on different chromosomes” appeared as >0.2, indicating these mapping statistics may have been affected by the prolonged storage and/or formalin fixation. These events were relatively rare and seem to be caused by either biological mutations occurring in small subclones in cancer tissue, or through storage-fixation artifacts, or both.

View this table:
Table II.

Correlation and coefficient of variables among duration of storage or fixation of samples with mapping statistics.

Interestingly, different mapping statistics were affected by the variations in storage and fixation duration. The storage length of FFPE samples affected the “Inward oriented pairs”, “Non-primary alignment”, and “Bases mapped (cigar)” (R2≥0.5) (Figure 1a and c). Meanwhile, the fixation length affected the “Unmapped reads” and “Singletons” (R2≥0.4) (Figure 1b and d). ΔΔCq shows similar but stronger pattern of correlations with mapping parameters to the storage length (Table II).

Figure 1.
Figure 1.

Correlation between storage and fixation length and mapping statistics. (a) Correlation between storage length and ratio of non-primary mapping (i.e. duplication). Vertical axis indicates the ratio of non-primary mapping reads against the total reads. Horizontal axis indicates the storage years of the FFPE samples. Dotted line indicates the linear regression. (b) Correlation between fixation length and ratio of unmapped reads. (c) Correlation between storage length and ratio of inward-oriented pairs. (d) Correlation between fixation length and ratio of singletons.

Storage and fixation length may not affect the detection of variants of major cancer mutations. Cancer profiling tests are mainly used to detect drug-targetable mutations that should be found in the majority of the sample analyzed. Considering the target sizes (~1.4 MB), around 660 germline variants per individual sample would be expected. Most of the variants found in this study were annotated as positive in 4.7 KJPN (93.7% to 99.8%), suggesting that most came from germline mutations. In addition, PANC-1 pancreatic cancer cells, which originated from a Caucasian patient, showed relatively high numbers (34 variants) of 4.7KJPN-negative variants. More than half of the samples showed variants in the COSMIC database but not in the 4.7KJPN. All of them (13 variants) are single nucleotide variants (SNVs) and detected in most of the tissues analyzed (Table III).

View this table:
Table III.

Cosmic mutations identified in this study.

Unexpectedly, we found a few dinucleotide variants (DNVs) per sample in our dataset. Sanger sequencing of the selected variants (SNVs and DNVs) was undertaken for 12 loci (Table IV and Figures 2 and 3). One locus, g.chr17: 29663624–29663625, is in the middle of the T stretch and A stretch inverted repeat (11 T followed by 11 A: chr17: 29663615–29552636). We failed to confirm the sequence of this locus for one sample and thus we could only verify 11 loci (Table III). All the DNVs except one were confirmed by the Sanger method and we speculated whether the two nucleotide changes occurred in one chromosome or were split in two chromosomes, as suggested by the database. Figure 2 indicates that the DNVs were basically localized in single reads, suggesting that most of the DNVs that appeared in this study should be considered as one variant rather than two SNVs (14). Two DNVs showed different minor allele frequencies between the first and second variant nucleotides (chr7:140185334–140185334 and chr9:13977498–13977499). In one of these, chr9:13977498– 13977499, we observed three different alleles in our collection. One allele corresponded to the reference, the DNV, and g.chr9: 13977499G>C SNV among our samples, indicating that the overlapped nucleotide between the DNV and SNV showed a larger minor allele frequency in the 4.7KJPN database. We could not identify overlapping SNVs for the other, chr7:140185334–140185334, in our dataset. We also verified indels by Sanger sequencing and agarose gel electrophoresis of the amplicons including the indel-detected regions. Figure 3 shows that all the four indels tested showed the expected one-base shift in the fluorogram (Figure 3a) or in the band sizes of the amplicon products (Figure 3b), showing the accuracy of the NCC oncopanel variant call data.

View this table:
Table IV.

Comparison between NCC oncopanel and Sanger sequencing.

Figure 2.
Figure 2.

Verification of DNVs found in the NCC oncopanel data by Sanger sequencing. Verification of dinucleotide variants (DNVs). Left image is generated with the Integrative Genomics Viewer (16). Right image is Sanger fluorogram corresponding to the DNV.

Figure 3.
Figure 3.

Verification of indels in NCC oncopanel data. (a) Verification of one nucleotide insertion by Sanger sequencing. Top image is generated with the Integrative Genomics Viewer and one base (Adenine) insertion is visible in the middle of the image. Bottom image is Sanger fluorogram corresponding to one-base insertion (b) Verification of indel variants. The genomic positions, amplicon sizes, and sample genotypes are indicated at the top of the gel image. The nucleotide lengths are indicated on the left and right sides of the gel image. The PCR templates are sample 14 (lanes 1, 3, and 7), sample 13 (lanes 2, 4, and 8), sample 8 (lane 5) and sample 10 (lane 6). The alt alleles are indicated with triangles in the gel image. “wt” and “alt” indicate the lengths of wild-type and of alternate alleles, respectively.

Discussion

The present study was focused on the robustness of a cancer profiling test for the quality of genomic DNA extracted from FFPE surgical specimens from various cancerous tissues with different storage conditions. To address this issue, we investigated a cancer profiling test, NCC oncopanel. Initial quality estimates by real-time PCR indicated that the quality of the extracted genomic DNA was not strongly correlated with the storage conditions. Similarly, the quality of the next generation sequencing reads was not different significantly, regardless of their storage conditions. However, the mapping quality of the sequence reads was affected by the FFPE storage conditions. Interestingly, the length of storage and fixation affected distinct mapping statistics of the NCC oncopanel cancer profiling test reads. Prolonged storage increased non-primary alignments and longer fixation caused an increase in unmapped reads. However, appropriate filtration of the variant calls produced reasonable results as an NCC oncopanel cancer profiling test, regardless of the variation in the storage or fixation length of FFPE samples. Our results indicate that the NCC oncopanel is a very robust cancer profiling test for most FFPE cancer specimens.

In the present study, we tested the quality of FFPE samples with real-time PCR and adjusted the amount of DNA sample based on the results. This may not be possible for most of the hospitals in Japan because the real-time PCR kits for the quality control of FFPE samples are not covered by Japanese health insurance. Therefore, judging the quality of FFPE samples for cancer profiling tests by storage periods or fixation length should be critical.

One of our interesting observations was that the mapping statistics can differ with different storage and fixation lengths. Increases in non-primary alignment resulted in increases in PCR duplication and may cause a decrease in SNV detection (15). Most likely, prolonged storage could cause an overall decrease in the “amplifiable DNA” in the FFPE samples and consequentially a decrease in the complexity of the NGS libraries. At the bedside, biopsy specimens are frequently used for cancer profiling tests. In the case of biopsy samples with prolonged storage, more sections may be required for cancer profiling tests than for newer biopsy samples. In the case of fixation, its effects are mainly to increase the sequence variations compared with the reference genome. C to T deamination is the mechanism of the increase in “unmapped reads” and “singletons” (5, 6). “Unmapped reads” and “singletons” would be caused by a few nucleotide changes in sequence reads and cause mismapping.

We identified a substantial number of DNVs that were not found in the genome variant databases of the general population. Wang et al. pointed out that there are three major mutational mechanisms for DNVs: combinations of independent SNVs, replication errors by DNA polymerase-zeta, and slippage at the repeat junctions (14). The first type of DNV mutation mechanism was detected in our dataset and showed different allele frequencies in the 4.7KJPN (Table IV). The third type of DNV was hg19/g.chr17:29663624-29663625, which we failed to verify in one sample because of the slippage during Sanger sequencing reactions. We believe some of the DNVs in Table IV are a result of the second mutational DNV mechanism, although the pattern of variation was not typical for polymerase-zeta (TC or GC>AA) (14). The variant databases did not have DNVs because the default option of the variant caller (GATK HaplotypeCaller) splits the MNVs into SNVs for subsequent joint genotyping in population studies.

The recommended conditions for FFPE samples for cancer profiling panel tests are as following: firstly, the prolonged stored samples (more than 4 years) may be detrimental for cancer profiling tests. Secondly, prolonged fixation (more than 3 days) should also be avoided. Ideally, the resected tumor samples should be cut for the better permeability of formalin. In summary, our results show that a cancer profiling test, the NCC oncopanel, is robust for FFPE sample quality. The storage and fixation conditions of samples tested in this study are within the expected range of FFPE samples at most Japanese hospitals. Hence, we conclude that most FFPE samples can be examined using cancer profiling tests.

Acknowledgements

The Authors thank Ms. Mika Takeuchi and Ms. Miyuki Ueki for their assistance with the sample preparation. They also thank H. Nikki March, Ph.D., from Edanz Group (https://en-author-services.edanzgroup.com/ac) for editing a draft of this manuscript. This work was supported by JSPS KAKENHI with the following grant numbers: JP19H01036 (to N. Tanuma), JP17K07187 (to H. Shima), JP19K08430 (Tamai), JP18K09363 (Mochizuki), and JP17K07193 (to J. Yasuda). This work was also supported by Foundation for Promotion of Cancer Research in Japan (Tamai) and Takeda Medical Foundation (Tamai).

Footnotes

  • Authors’ Contributions

    J.Y. and I.S. planned the study. S.I and T.M. performed the experiments. M.M., K.Y., and K.T. helped conduct the experimental procedures. Bioinformatics analysis was performed by S.I., N.T., and J.Y. The manuscript was written by S.I. and J.Y. H.S. contributed to the general management of the project.

  • Conflicts of Interest

    There are no conflicts of interest to disclose.

  • Received December 28, 2020.
  • Revision received January 27, 2021.
  • Accepted February 1, 2021.

References

  1. ↵
    1. Hellmann MD,
    2. Ciuleanu TE,
    3. Pluzanski A,
    4. Lee JS,
    5. Otterson GA,
    6. Audigier-Valette C,
    7. Minenza E,
    8. Linardou H,
    9. Burgers S,
    10. Salman P,
    11. Borghaei H,
    12. Ramalingam SS,
    13. Brahmer J,
    14. Reck M,
    15. O’Byrne KJ,
    16. Geese WJ,
    17. Green G,
    18. Chang H,
    19. Szustakowski J,
    20. Bhagavatheeswaran P,
    21. Healey D,
    22. Fu Y,
    23. Nathan F and
    24. Paz-Ares L
    : Nivolumab plus ipilimumab in lung cancer with a high tumor mutational burden. N Engl J Med 378(22): 2093-2104, 2018. PMID: 29658845. DOI: 10.1056/NEJMoa1801946
    OpenUrlCrossRefPubMed
  2. ↵
    1. Sunami K,
    2. Ichikawa H,
    3. Kubo T,
    4. Kato M,
    5. Fujiwara Y,
    6. Shimomura A,
    7. Koyama T,
    8. Kakishima H,
    9. Kitami M,
    10. Matsushita H,
    11. Furukawa E,
    12. Narushima D,
    13. Nagai M,
    14. Taniguchi H,
    15. Motoi N,
    16. Sekine S,
    17. Maeshima A,
    18. Mori T,
    19. Watanabe R,
    20. Yoshida M,
    21. Yoshida A,
    22. Yoshida H,
    23. Satomi K,
    24. Sukeda A,
    25. Hashimoto T,
    26. Shimizu T,
    27. Iwasa S,
    28. Yonemori K,
    29. Kato K,
    30. Morizane C,
    31. Ogawa C,
    32. Tanabe N,
    33. Sugano K,
    34. Hiraoka N,
    35. Tamura K,
    36. Yoshida T,
    37. Fujiwara Y,
    38. Ochiai A,
    39. Yamamoto N and
    40. Kohno T
    : Feasibility and utility of a panel testing for 114 cancer-associated genes in a clinical setting: A hospital-based study. Cancer Sci 110(4): 1480-1490, 2019. PMID: 30742731. DOI: 10.1111/cas.13969
    OpenUrlCrossRefPubMed
  3. ↵
    1. Srinivasan M,
    2. Sedmak D and
    3. Jewell S
    : Effect of fixatives and tissue processing on the content and integrity of nucleic acids. Am J Pathol 161(6): 1961-1971, 2002. PMID: 12466110. DOI: 10.1016/S0002-9440(10)64472-0
    OpenUrlCrossRefPubMed
  4. ↵
    1. Vitošević K,
    2. Todorović M,
    3. Varljen T,
    4. Slović Ž,
    5. Matić S and
    6. Todorović D
    : Effect of formalin fixation on pcr amplification of DNA isolated from healthy autopsy tissues. Acta Histochem 120(8): 780-788, 2018. PMID: 30224244. DOI: 10.1016/j.acthis.2018.09.005
    OpenUrlCrossRefPubMed
  5. ↵
    1. Do H and
    2. Dobrovic A
    : Dramatic reduction of sequence artefacts from DNA isolated from formalin-fixed cancer biopsies by treatment with uracil- DNA glycosylase. Oncotarget 3(5): 546-558, 2012. PMID: 22643842. DOI: 10.18632/oncotarget.503
    OpenUrlCrossRefPubMed
  6. ↵
    1. Prentice LM,
    2. Miller RR,
    3. Knaggs J,
    4. Mazloomian A,
    5. Aguirre Hernandez R,
    6. Franchini P,
    7. Parsa K,
    8. Tessier-Cloutier B,
    9. Lapuk A,
    10. Huntsman D,
    11. Schaeffer DF and
    12. Sheffield BS
    : Formalin fixation increases deamination mutation signature but should not lead to false positive mutations in clinical practice. PLoS One 13(4): e0196434, 2018. PMID: 29698444. DOI: 10.1371/journal.pone.0196434
    OpenUrlCrossRefPubMed
  7. ↵
    1. Bolger AM,
    2. Lohse M and
    3. Usadel B
    : Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics 30(15): 2114-2120, 2014. PMID: 24695404. DOI: 10.1093/bioinformatics/btu170
    OpenUrlCrossRefPubMed
  8. ↵
    1. Li H
    : Aligning sequence reads, clone sequences and assembly contigs with bwa-mem. arXive: 1303.3997v1301 [q-bio.GN], 2013.
  9. ↵
    1. Li H,
    2. Handsaker B,
    3. Wysoker A,
    4. Fennell T,
    5. Ruan J,
    6. Homer N,
    7. Marth G,
    8. Abecasis G,
    9. Durbin R and 1000 Genome Project Data Processing Subgroup
    : The sequence alignment/map format and SAMtools. Bioinformatics 25(16): 2078-2079, 2009. PMID: 19505943. DOI: 10.1093/bioinformatics/btp352
    OpenUrlCrossRefPubMed
  10. ↵
    1. Wang K,
    2. Li M and
    3. Hakonarson H
    : ANNOVAR: functional annotation of genetic variants from high-throughput sequencing data. Nucleic Acids Res 38(16): e164, 2010. PMID: 20601685. DOI: 10.1093/nar/gkq603
    OpenUrlCrossRefPubMed
  11. ↵
    1. Tate JG,
    2. Bamford S,
    3. Jubb HC,
    4. Sondka Z,
    5. Beare DM,
    6. Bindal N,
    7. Boutselakis H,
    8. Cole CG,
    9. Creatore C,
    10. Dawson E,
    11. Fish P,
    12. Harsha B,
    13. Hathaway C,
    14. Jupe SC,
    15. Kok CY,
    16. Noble K,
    17. Ponting L,
    18. Ramshaw CC,
    19. Rye CE,
    20. Speedy HE,
    21. Stefancsik R,
    22. Thompson SL,
    23. Wang S,
    24. Ward S,
    25. Campbell PJ and
    26. Forbes SA
    : COSMIC: the catalogue of somatic mutations in cancer. Nucleic Acids Res 47(D1): D941-D947, 2019. PMID: 30371878. DOI: 10.1093/nar/gky1015
    OpenUrlCrossRefPubMed
  12. ↵
    1. Tadaka S,
    2. Saigusa D,
    3. Motoike IN,
    4. Inoue J,
    5. Aoki Y,
    6. Shirota M,
    7. Koshiba S,
    8. Yamamoto M and
    9. Kinoshita K
    : jMorp: Japanese multi omics reference panel. Nucleic Acids Res 46(D1): D551-D557, 2018. PMID: 29069501. DOI: 10.1093/nar/gkx978
    OpenUrlCrossRefPubMed
  13. ↵
    1. Li H
    : A statistical framework for SNP calling, mutation discovery, association mapping and population genetical parameter estimation from sequencing data. Bioinformatics 27(21): 2987-2993, 2011. PMID: 21903627. DOI: 10.1093/bioinformatics/btr509
    OpenUrlCrossRefPubMed
  14. ↵
    1. Wang Q,
    2. Pierce-Hoffman E,
    3. Cummings BB,
    4. Alföldi J,
    5. Francioli LC,
    6. Gauthier LD,
    7. Hill AJ,
    8. O’Donnell-Luria AH, Genome Aggregation Database Production Team, Genome Aggregation Database Consortium,
    9. Karczewski KJ and
    10. MacArthur DG
    : Landscape of multi-nucleotide variants in 125,748 human exomes and 15,708 genomes. Nat Commun 11(1): 2539, 2020. PMID: 32461613. DOI: 10.1038/s41467-019-12438-5
    OpenUrlCrossRefPubMed
  15. ↵
    1. Liu Q,
    2. Guo Y,
    3. Li J,
    4. Long J,
    5. Zhang B and
    6. Shyr Y
    : Steps to ensure accuracy in genotype and SNP calling from Illumina sequencing data. BMC Genomics 13 Suppl 8: S8, 2012. PMID: 23281772. DOI: 10.1186/1471-2164-13-S8-S8
    OpenUrlCrossRef
  16. ↵
    1. Robinson JT,
    2. Thorvaldsdóttir H,
    3. Winckler W,
    4. Guttman M,
    5. Lander ES,
    6. Getz G and
    7. Mesirov JP
    : Integrative genomics viewer. Nat Biotechnol 29(1): 24-26, 2011. PMID: 21221095. DOI: 10.1038/nbt.1754
    OpenUrlCrossRefPubMed
Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools
Reprints and Permissions
Robustness of a Cancer Profiling Test Using Formalin-fixed Paraffin Embedded Tumor Specimens
SHIN ITO, IKURO SATO, MAI MOCHIZUKI, KAZUNORI YAMAGUCHI, KEIICHI TAMAI, TAKAMICHI MINATO, NOBUHIRO TANUMA, HIROSHI SHIMA, JUN YASUDA
Anticancer Research Mar 2021, 41 (3) 1341-1348; DOI: 10.21873/anticanres.14891
Twitter logo Facebook logo Mendeley logo

Jump to section

Keywords