Median (minimum-maximum) Total number of mapped bases 2,504,424,995 (718,875,018-6,059,712,422) Read depth of the target region a Mean: 49.8 (13.9-117.3) Number of target sites with both normal and tumor depths ≥ 10 27,516,112 (18,965,079-29,556,816) Number of tumor-specific substitutions and small indels in the target 47 (1-1,059) Supplementary Table S1. Sequenced reads per tumor and normal sample. a The target region indicates the genomic region of the exome-capturing libraries.
15
Embed
et al - Clinical Cancer Research · USAF1 ZMYND10 Supplementary Table S6. Comparison of significantly mutated genes and previously reported cancer driver genes. a c, , et al (((((,
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Median (minimum-maximum)Total number of mapped bases 2,504,424,995 (718,875,018-6,059,712,422)Read depth of the target regiona Mean: 49.8 (13.9-117.3)Number of target sites with both normal andtumor depths ≥ 10 27,516,112 (18,965,079-29,556,816)
Number of tumor-specific substitutions andsmall indels in the target 47 (1-1,059)
Supplementary Table S1. Sequenced reads per tumor and normal sample.
aThe target region indicates the genomic region of the exome-capturing libraries.
Comparison of significantly mutated genes detected using various MuTect thresholds."Default" used the recommended default threshold (Cibulskis et al . 2013), while "Ourpipeline" used a more stringent threshold (see Supplementary Methods). Potential truepositives include previously reported driver genes (45 genes included in SupplementaryTable S6). Potential false positives include olfactory receptor genes and other known falsepositives (mucins, CSMD1 , CSMD3 , TTN , RYR2 , and RYR3 ).
Statistics are shown for the nine substitution types that were considered. The substitution types are strand-collapsed. For instance, A→C indicates A to C substitutions on either the top or bottom of the chromosome strand.
Supplementary Table S3. Comparison of the number and rate of substitutions between never-smokers and smokers.
Never-smokers Smokers
Factors consideredfor BMR calculation
Minimum number of SN sitesrequired for local BMR
calculation (ΣNSN)
Number ofdetected genes
Number ofdetected potentialtrue positives (%)
Number ofdetected potentialfalse positives (%)
From previous approaches Base type N/A 90 9 (10%) 9 (10%)
Our approach, min. ΣNSN=500Base type,local BMR,
site degeneracy500 74 6 (8%) 1 (1%)
Our approach, min. ΣNSN=3000Base type,local BMR,
site degeneracy3000 48 6 (13%) 2 (4%)
Our approach, min. ΣNSN=5000Base type,local BMR,
site degeneracy5000 36 6 (17%) 2 (6%)
Supplementary Table S4. Comparison of various methods used to compute the background mutation rate (BMR).
Comparison of various methods used to compute the background mutation rate (BMR). Previous studies (Ding et al .2008, Kan et al . 2010) calculated the BMR after taking into account base type, but not local BMR variations. On the otherhand, we took both into account and also considered site degeneracy when computing BMR (see SupplementaryMethods). We used various parameters (minimum number of synonymous [SN] sites = ΣNSN) in the local BMRcalculations and selected the best performing option: ΣNSN = 3000. Potential true positives include previously reporteddriver genes (45 genes included in Supplementary Table S6). Potential false positives include olfactory receptor genes andother known false positives (mucins, CSMD1 , CSMD3 , TTN , RYR2 , and RYR3 ).
Gene symbol RefSeq ID q value Prevalence Number of affectedpatients
Forty-eight genes passed the Poisson tests after correcting for multiple tests. Genes are sorted by q value. Thetop (+) or bottom (-) strand on which the reference cDNA is transcribed is indicated. “ORF length” indicates theORF length (bp) of the reference cDNA. The number of mutations followed by the mutation type discovered in thegene is shown. The listed mutation types include: S, synonymous; N, nonsynonymous; F, 5' splice-site; T, 3' splice-site; *, stop gain; fp, frame-preserving indel; fs, frame-shifting indel; D, deletion; I, insertion. Otherwise, as in Table2.
Supplementary Table S6. Comparison of significantly mutated genes and previously reported cancer driver genes.
aPotential cancer driver genes reported in previous studies (see Supplementary Table S15).bSignificantly mutated genes discovered in our study.cGenes detected as significantly mutated from in studies.dNovel potential cancer driver genes that were only found in our study.ePreviously reported cancer driver genes that demonstrated ≥ 1 detected mutation in this study but whose level ofsignificance did not pass our cutoff value.
Our studyb Found in bothc Novel candidatesd Driver candidates with ≥1detected mutation(s)e
CNV region Chromosomallocation
Amplification orDeletion GISTIC q value Genes located in CNV region
Supplementary Table S7. Thirteen significantly altered CNV regions detected in our discovery cohort.CNV regions are sorted by the q values determined using GISTIC (see Supplementary Methods). The representativegene of each CNV region is shown in bold.
Gene symbol Location Type Drugs
COL11A1 a Extracellular space Collagen Collagenase clostridium hystolyticumCOL6A3 a Extracellular space Collagen Collagenase clostridium hystolyticumEGFR b Plasma membrane Kinase Erlotinib/gemcitabineERBB2 c Plasma membrane Kinase TrastuzumabF8 d Extracellular space Peptidase Drotrecogin alfaPIK3CA e Cytoplasm Kinase NVP-BEZ235SQLE f Cytoplasm Enzyme TerbinafineTERT g Nucleus Enzyme GRN163L
Four oncogenes (EGFR , ERBB2 , PIK3CA , and TERT ) are known targets of available drugs. Proteins expressedby two collagen genes (COL11A1 and COL6A3 ) could be degraded by collagenases. Coagulation factor F8could be inhibited by drotrecogin alfa. SQLE (squalene monooxygenase) could be inhibited by terbinafine.
Supplementary Table S8. Druggable target candidates in the 22 significantly mutated genes and the 60 genes inthe significantly altered CNV regions detected in the discovery cohort.
aMandl et al ., J. Clin. Invest., 1953, 32:1323-9.bByrne and Garst, Curr. Oncol. Rep., 2005, 7:241-7.cStephens et al ., Nature, 2004, 431:525-6.dRanieri et al ., N. Engl. J. Med., 2012, 366:2055-64.eEngelman et al ., Nat. Med., 2008, 14:1351-6.fNowosielski et al ., J. Chem. Inf. Model., 2011, 51:455-62.gAsai et al ., Cancer Res., 2003, 63:3931-9.
Pathway Referencedatabase P value Benjamini-Hochberg
Thyroid cancer KEGG_PATHWAY 8.7×10-3 4.2×10-2 TP53, KRAS, MYC
Supplementary Table S9. Pathways significantly enriched in the 22 significantly mutated genes and the 60 genes included in the significantly altered CNV regions that were detected in the discovery cohort.
Supplementary Table S10.Cox multivariable regression analysis for RB pathway alterations.
Parametera HR 95% CI P HR 95% CI P
(trohoc yrevocsiD n ()031 = n = 27)429.0)001> ot 0(0969.0)42.2 ot 64.0(20.1elam susrev elameF
Age, ≥60 versus <60 0.87 (0.49 to 1.55) 0.636 1.84 (0.69 to 4.92) 0.224Smoker versus never-smoker 1.21 (0.56 to 2.65) 0.626 0 (0 to >100) 0.924Tumor status, T2/T3/T4 versus T1 1.67 (0.96 to 2.91) 0.07 >100 (0 to >100) 0.906Nodal status, N1/N2 versus N0 1.47 (0.78 to 2.75) 0.234 2.51 (0.41 to 15.48) 0.321LVI, positive versus negative 2.59 (1.42 to 4.74) 0.002 2.69 (0.82 to 8.86) 0.104Chemotherapy, treated versus untreated 0.81 (0.41 to 1.6) 0.536 0.69 (0.2 to 2.35) 0.551RB pathway, altered versus unaltered 2.26 (1.29 to 3.98) 0.005 5.14 (1.06 to 25) 0.043
(trohoc noitadilaV n ()95 = n = 13)729.0)001> ot 0(001>791.0)71.53 ot 84.0(11.4elam susrev elameF
Age, ≥60 versus <60 0.4 (0.16 to 0.98) 0.044 0 (0 to >100) 0.898Smoker versus never-smoker 2.87 (0.33 to 25.23) 0.341 >100 (0 to >100) 0.954Tumor status, T2/T3/T4 versus T1 0.79 (0.28 to 2.24) 0.659 0 (0 to >100) 0.901Nodal status, N1/N2 versus N0 1.12 (0.33 to 3.74) 0.86 0.38 (0.02 to 8.08) 0.532LVI, positive versus negative 2.42 (0.87 to 6.69) 0.089 >100 (0 to >100) 0.905Chemotherapy, treated versus untreated 2.54 (0.84 to 7.72) 0.099 0 (0 to >100) 0.902RB pathway, altered versus unaltered 5.56 (1.52 to 20.28) 0.009 >100 (0 to >100) 0.945Abbreviations: LVI, lymphovascular invasion; HR, hazard ratio; CI, confidence interval.aCovariates include alterations in the RB pathway and various clinical data such as sex, age, and smoking status.
Early stage (stages I & II) Late stage (stages III & IV)b
bThere is a high possibility that the number of patients in this section is insufficient for an accurate multivariable Cox hazardmodel, and, therefore, the results should be cautiously interpreted (Peduzzi P et al., Journal of clinical epidemiology, 1995,48:1503-10).
Supplementary Table S11.Cox univariable regression analysis of RB pathway alterations.
Parametera HR 95% CI P HR 95% CI P
Discovery cohort (n ()031 = n = 27)447.0)88.2 ot 74.0(61.1735.0)24.1 ot 15.0(58.0elam susrev elameF
Age, ≥60 versus <60 0.92 (0.55 to 1.54) 0.747 1.12 (0.45 to 2.81) 0.804Smoker versus never-smoker 1.26 (0.76 to 2.09) 0.371 0.64 (0.26 to 1.6) 0.34Tumor status, T2/T3/T4 versus T1 1.39 (0.82 to 2.35) 0.222 24.3 (0.02 to >100) 0.393Nodal status, N1/N2 versus N0 2.07 (1.15 to 3.73) 0.015 1.72 (0.39 to 7.49) 0.473LVI, positive versus negative 2.78 (1.62 to 4.78) <0.001 1.8 (0.66 to 4.92) 0.25Chemotherapy, treated versus untreated 1.36 (0.74 to 2.48) 0.32 1.15 (0.41 to 3.23) 0.794RB pathway, altered versus unaltered 2.38 (1.38 to 4.11) 0.002 2.55 (0.77 to 8.44) 0.127
Validation cohort (n ()95 = n = 13)372.0)38.82 ot 93.0(43.3102.0)33.4 ot 37.0(87.1elam susrev elameF
Age, ≥60 versus <60 0.48 (0.2 to 1.16) 0.105 0.09 (0.01 to 0.81) 0.032Smoker versus never-smoker 0.89 (0.37 to 2.16) 0.803 0.48 (0.06 to 4.17) 0.507Tumor status, T2/T3/T4 versus T1 1.29 (0.49 to 3.35) 0.606 0.57 (0.1 to 3.15) 0.519Nodal status, N1/N2 versus N0 1.39 (0.46 to 4.16) 0.558 1.56 (0.28 to 8.58) 0.61LVI, positive versus negative 1.76 (0.68 to 4.61) 0.247 0.94 (0.11 to 8.27) 0.957Chemotherapy, treated versus untreated 1.67 (0.68 to 4.07) 0.264 0.75 (0.09 to 6.49) 0.794RB pathway, altered versus unaltered 3.26 (1.17 to 9.12) 0.024 0.04 (0 to >100) 0.562Abbreviations: LVI, lymphovascular invasion; HR, hazard ratio; CI, confidence interval.aCovariates include alterations in the RB pathway and various clinical data such as sex, age, and smoking status.bThere is a high possibility that the number of patients in this section is insufficient for an accurate univariable Cox hazardmodel, and, therefore, the results should be cautiously interpreted (Peduzzi P et al., Journal of clinical epidemiology, 1995,48:1503-10).
Early stage (stages I & II) Late stage (stages III & IV)b
Supplementary Table S12. Comparison of mutations detected using whole-exome sequencing(WES) and Sequenom in selected genes (EGFR , KRAS , PIK3CA , and BRAF ).
The sensitivity of WES was computed (WES sensitivity) after assuming that the resultsdetermined using Sequenom were true positives. Detected: substitutions detected using bothWES and Sequenom; Sequenom-specific: substitutions detected using only Sequenom; WES-specific: substitutions detected using only WES.