In testing differentially expressed genes between tumor and healthy tissues, data are usually collected in paired form. However, incomplete paired data often occur. While extensive statistical researches exist for paired data with incompleteness in both arms, hardly any recent work can be found on paired data with incompleteness in single arm. This paper aims to fill this gap by proposing some new methods, namely, P-value pooling methods and a nonparametric combination test. Simulation studies are conducted to investigate the performance of the proposed methods in terms of type I error and power at small to moderate sample sizes. A real data set from The Cancer Genome Atlas (TCGA) breast cancer study is analyzed using the proposed methods.
WangZGersteinMSnyderM. RNA-Seq: a revolutionary tool for transcriptomics. Nat Rev Genet2009; 10: 57–63.
2.
MortazaviAWilliamsBAMcCueKet al.Mapping and quantifying mammalian transcriptomes by RNA-Seq. Nat Meth2008; 5: 621–628.
3.
DaveyJWHohenlohePAEtterPDet al.Genome-wide genetic marker discovery and genotyping using next-generation sequencing. Nat Rev Genet2011; 12: 499–510.
4.
WilsonCARamosLVillaseorMRet al.Localization of human BRCA1 and its loss in high-grade, non-inherited breast carcinomas. Nat Genet1999; 21: 236–240.
5.
KuriakoseMAChenWTHeZMet al.Selection and validation of differentially expressed genes in head and neck cancer. Cell Mol Life Sci2004; 61: 1372–1383.
6.
LapointeJLiCHigginsJPet al.Gene expression profiling identifies clinically relevant subtypes of prostate cancer. Proc Natl Acad Sci U S A2004; 101: 811–816.
7.
FengQHawesSESternJEet al.DNA methylation in tumor and matched normal tissues from non-small cell lung cancer patients. Cancer Epidemiol Biomarker Prevent2008; 17: 645–654.
8.
FinakGBertosNPepinFet al.Stromal gene expression predicts clinical outcome in breast cancer. Nat Med2008; 14: 518–527.
9.
KimBSKimILeeSet al.Statistical methods of translating microarray data into clinically relevant diagnostic information in colorectal cancer. Bioinformatics2005; 21: 517–528.
10.
LinPEStiversLE. On difference of means with incomplete data. Biometrika1974; 61: 325–334.
11.
EkbohmG. On comparing means in the paired case with incomplete data on both responses. Biometrika1976; 63: 299–304.
12.
BhojDS. Testing equality of means of correlated variates with missing observations on both responses. Biometrika1978; 65: 225–228.
13.
KuanPFHuangB. A simple and robust method for partially matched samples using the pvalues pooling approach. Stat Med2013; 32: 3247–3259.
14.
SamawiHMYuLVogelR. On some nonparametric tests for partially observed correlated data: Proposing new tests. J Stat Theory Appl2015; 14: 131–155.
15.
RempalaGALooneySW. Asymptotic properties of a two sample randomized test for partially dependent data. J Stat Plan Inference2006; 136: 68–89.
16.
YuDLimJLiangFet al.Permutation test for incomplete paired data with application to cDNA microarray data. Computat Stat Data Analysis2012; 56: 510–521.
17.
AmroLPaulyM. Permuting incomplete paired data: a novel exact and asymptotic correct randomization test. J Stat Comput Simul2017; 87: 1148–1159.
18.
GuoBYuanY. A comparative review of methods for comparing means using partially paired data. Stat Meth Med Res2017; 26: 1323–1340.
19.
HuangXSternDFZhaoH. Transcriptional profiles from paired normal samples offer complementary information on cancer patient survival evidence from TCGA pan-cancer data. Scientific Reports2016; 6: 20567–20567.
RitchieMEPhipsonBWuDet al.Limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucl Acids Res2015; 43: e47–e47.
22.
LoveMIHuberWAndersS. Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol2014; 15: 550–550.
23.
MehtaJSGurlandJ. Testing equality of means in the presence of correlation. Biometrika1969; 56: 119–126.
24.
LinPE. Procedures for testing the difference of means with incomplete data. J Am Stat Assoc1973; 68: 699–703.
25.
MorrisonDF. A test for equality of means of correlated variates with missing data on one response. Biometrika1973; 60: 101–105.
26.
LittleRJA. Inference about means from incomplete multivariate data. Biometrika1976; 63: 593–604.
27.
EkbohmG. Comparing means in the paired case with missing data on one response. Biometrika1976; 63: 169–172.
28.
MehtaJSGurlandJ. A test for equality of means in the presence of correlation and missing values. Biometrika1973; 60: 211–213.
29.
LinPEStiversLE. Testing for equality of means with incomplete data on one variable: a Monte Carlo study. J Am Stat Assoc1975; 70: 190–193.
30.
DeSantisCMaJBryanLet al.Breast cancer statistics, 2013. CA: Cancer J Clinicians2014; 64: 52–62.
31.
FordDEastonDFStrattonMet al.Genetic heterogeneity and penetrance analysis of the BRCA1 and BRCA2 genes in breast cancer families. Am J Human Genet1998; 62: 676–689.
32.
RahmanNSealSThompsonDet al.PALB2, which encodes a BRCA2-interacting protein, is a breast cancer susceptibility gene. Nat Genet2007; 39: 165–167.
33.
HarariDYardenY. Molecular mechanisms underlying ErbB2/HER2 action in breast cancer. Oncogene2000; 19: 6102–6114.
34.
MunozMHendersonMHaberMet al.Role of the MRP1/ABCC1 multidrug transporter protein in cancer. IUBMB Life2007; 59: 752–757.
35.
De JongMMNolteIMTe MeermanGJet al.Genes other than BRCA1 and BRCA2 involved in breast cancer susceptibility. J Med Genet2002; 39: 225–242.
36.
RosenthalR. Combining results of independent studies. Psychol Bull1978; 85: 185–193.
37.
BaileyTLGribskovM. Combining evidence using p-values: application to sequence homology searches. Bioinformatics1998; 14: 48–54.
38.
HartungJ. A note on combining dependent tests of significance. Biometric J1999; 41: 849–855.
39.
BrownMB. A method for combining non-independent, one-sided tests of significance. Biometrics1975; 31: 987–992.
40.
KostJTMcDermottMP. Combining dependent p-values. Stat Probabil Lett2002; 60: 183–190.
41.
HouCD. A simple approximation for the distribution of the weighted combination of non-independent or independent probabilities. Stat Probab Lett2005; 73: 179–187.
42.
ZaykinDV. Optimally weighted Z-test is a powerful method for combining probabilities in meta-analysis. J Evolution Biol2011; 24: 1836–1841.
