Correlation between miRseq expression and clinical features
Uterine Corpus Endometrioid Carcinoma (Primary solid tumor)
15 January 2014  |  analyses__2014_01_15
Maintainer Information
Citation Information
doi:10.7908/C1ZG6QQZ
Maintained by Juok Cho (Broad Institute)
Cite as Broad Institute TCGA Genome Data Analysis Center (2014): Correlation between miRseq expression and clinical features. Broad Institute of MIT and Harvard. doi:10.7908/C1ZG6QQZ
Overview
Introduction

This pipeline uses various statistical tests to identify miRs whose expression levels correlated to selected clinical features.

Summary

Testing the association between 555 miRs and 5 clinical features across 471 samples, statistically thresholded by Q value < 0.05, 4 clinical features related to at least one miRs.

  • 5 miRs correlated to 'Time to Death'.

    • HSA-MIR-497 ,  HSA-LET-7G ,  HSA-MIR-628 ,  HSA-MIR-195 ,  HSA-MIR-34A

  • 45 miRs correlated to 'AGE'.

    • HSA-MIR-424 ,  HSA-MIR-1247 ,  HSA-MIR-337 ,  HSA-MIR-935 ,  HSA-MIR-199A-1 ,  ...

  • 102 miRs correlated to 'HISTOLOGICAL.TYPE'.

    • HSA-MIR-9-3 ,  HSA-MIR-9-2 ,  HSA-MIR-9-1 ,  HSA-MIR-934 ,  HSA-MIR-34A ,  ...

  • 9 miRs correlated to 'RADIATIONS.RADIATION.REGIMENINDICATION'.

    • HSA-MIR-3613 ,  HSA-MIR-128-1 ,  HSA-MIR-128-2 ,  HSA-MIR-628 ,  HSA-MIR-107 ,  ...

  • No miRs correlated to 'COMPLETENESS.OF.RESECTION'

Results
Overview of the results

Complete statistical result table is provided in Supplement Table 1

Table 1.  Get Full Table This table shows the clinical features, statistical methods used, and the number of miRs that are significantly associated with each clinical feature at Q value < 0.05.

Clinical feature Statistical test Significant miRs Associated with                 Associated with
Time to Death Cox regression test N=5 shorter survival N=0 longer survival N=5
AGE Spearman correlation test N=45 older N=4 younger N=41
HISTOLOGICAL TYPE ANOVA test N=102        
RADIATIONS RADIATION REGIMENINDICATION t test N=9 yes N=6 no N=3
COMPLETENESS OF RESECTION ANOVA test   N=0        
Clinical variable #1: 'Time to Death'

5 miRs related to 'Time to Death'.

Table S1.  Basic characteristics of clinical feature: 'Time to Death'

Time to Death Duration (Months) 0-191.8 (median=20.9)
  censored N = 414
  death N = 54
     
  Significant markers N = 5
  associated with shorter survival 0
  associated with longer survival 5
List of 5 miRs significantly associated with 'Time to Death' by Cox regression test

Table S2.  Get Full Table List of 5 miRs significantly associated with 'Time to Death' by Cox regression test

HazardRatio Wald_P Q C_index
HSA-MIR-497 0.66 1.066e-05 0.0059 0.334
HSA-LET-7G 0.51 1.425e-05 0.0079 0.333
HSA-MIR-628 0.68 5.007e-05 0.028 0.357
HSA-MIR-195 0.69 5.752e-05 0.032 0.346
HSA-MIR-34A 0.72 6.424e-05 0.035 0.365

Figure S1.  Get High-res Image As an example, this figure shows the association of HSA-MIR-497 to 'Time to Death'. four curves present the cumulative survival rates of 4 quartile subsets of patients. P value = 1.07e-05 with univariate Cox regression analysis using continuous log-2 expression values.

Clinical variable #2: 'AGE'

45 miRs related to 'AGE'.

Table S3.  Basic characteristics of clinical feature: 'AGE'

AGE Mean (SD) 63.69 (11)
  Significant markers N = 45
  pos. correlated 4
  neg. correlated 41
List of top 10 miRs significantly correlated to 'AGE' by Spearman correlation test

Table S4.  Get Full Table List of top 10 miRs significantly correlated to 'AGE' by Spearman correlation test

SpearmanCorr corrP Q
HSA-MIR-424 -0.3239 6.121e-13 3.4e-10
HSA-MIR-1247 -0.2681 4.08e-09 2.26e-06
HSA-MIR-337 -0.2474 5.501e-08 3.04e-05
HSA-MIR-935 0.2683 7.169e-08 3.96e-05
HSA-MIR-199A-1 -0.2381 1.762e-07 9.71e-05
HSA-MIR-409 -0.2355 2.422e-07 0.000133
HSA-MIR-516A-1 0.2998 2.63e-07 0.000144
HSA-MIR-214 -0.2348 2.711e-07 0.000149
HSA-MIR-199A-2 -0.2343 2.787e-07 0.000152
HSA-MIR-134 -0.2342 2.825e-07 0.000154

Figure S2.  Get High-res Image As an example, this figure shows the association of HSA-MIR-424 to 'AGE'. P value = 6.12e-13 with Spearman correlation analysis. The straight line presents the best linear regression.

Clinical variable #3: 'HISTOLOGICAL.TYPE'

102 miRs related to 'HISTOLOGICAL.TYPE'.

Table S5.  Basic characteristics of clinical feature: 'HISTOLOGICAL.TYPE'

HISTOLOGICAL.TYPE Labels N
  ENDOMETRIOID ENDOMETRIAL ADENOCARCINOMA 363
  MIXED SEROUS AND ENDOMETRIOID 19
  SEROUS ENDOMETRIAL ADENOCARCINOMA 89
     
  Significant markers N = 102
List of top 10 miRs differentially expressed by 'HISTOLOGICAL.TYPE'

Table S6.  Get Full Table List of top 10 miRs differentially expressed by 'HISTOLOGICAL.TYPE'

ANOVA_P Q
HSA-MIR-9-3 2.082e-32 1.16e-29
HSA-MIR-9-2 2.824e-29 1.56e-26
HSA-MIR-9-1 3.332e-29 1.84e-26
HSA-MIR-934 3.206e-25 1.77e-22
HSA-MIR-34A 1.49e-21 8.21e-19
HSA-MIR-375 7.766e-20 4.27e-17
HSA-MIR-221 4.451e-19 2.44e-16
HSA-MIR-195 4.6e-19 2.52e-16
HSA-MIR-452 1.16e-18 6.34e-16
HSA-MIR-190B 2.48e-18 1.35e-15

Figure S3.  Get High-res Image As an example, this figure shows the association of HSA-MIR-9-3 to 'HISTOLOGICAL.TYPE'. P value = 2.08e-32 with ANOVA analysis.

Clinical variable #4: 'RADIATIONS.RADIATION.REGIMENINDICATION'

9 miRs related to 'RADIATIONS.RADIATION.REGIMENINDICATION'.

Table S7.  Basic characteristics of clinical feature: 'RADIATIONS.RADIATION.REGIMENINDICATION'

RADIATIONS.RADIATION.REGIMENINDICATION Labels N
  NO 131
  YES 340
     
  Significant markers N = 9
  Higher in YES 6
  Higher in NO 3
List of 9 miRs differentially expressed by 'RADIATIONS.RADIATION.REGIMENINDICATION'

Table S8.  Get Full Table List of 9 miRs differentially expressed by 'RADIATIONS.RADIATION.REGIMENINDICATION'

T(pos if higher in 'YES') ttestP Q AUC
HSA-MIR-3613 -5.17 4.481e-07 0.000249 0.6534
HSA-MIR-128-1 4.93 1.382e-06 0.000766 0.64
HSA-MIR-128-2 4.69 4.207e-06 0.00233 0.6306
HSA-MIR-628 -4.6 6.666e-06 0.00368 0.6289
HSA-MIR-107 4.57 7.194e-06 0.00396 0.6198
HSA-MIR-361 4.39 1.635e-05 0.00899 0.6199
HSA-MIR-181D 4.31 2.327e-05 0.0128 0.6287
HSA-MIR-146A -4.16 4.37e-05 0.0239 0.6163
HSA-MIR-103-1 4.08 6.137e-05 0.0336 0.6187

Figure S4.  Get High-res Image As an example, this figure shows the association of HSA-MIR-3613 to 'RADIATIONS.RADIATION.REGIMENINDICATION'. P value = 4.48e-07 with T-test analysis.

Clinical variable #5: 'COMPLETENESS.OF.RESECTION'

No miR related to 'COMPLETENESS.OF.RESECTION'.

Table S9.  Basic characteristics of clinical feature: 'COMPLETENESS.OF.RESECTION'

COMPLETENESS.OF.RESECTION Labels N
  R0 323
  R1 24
  R2 15
  RX 27
     
  Significant markers N = 0
Methods & Data
Input

  • Expresson data file = UCEC-TP.miRseq_RPKM_log2.txt

  • Clinical data file = UCEC-TP.merged_data.txt

  • Number of patients = 471

  • Number of miRs = 555

  • Number of clinical features = 5

Survival analysis

For survival clinical features, Wald's test in univariate Cox regression analysis with proportional hazards model (Andersen and Gill 1982) was used to estimate the P values using the 'coxph' function in R. Kaplan-Meier survival curves were plot using the four quartile subgroups of patients based on expression levels

Correlation analysis

For continuous numerical clinical features, Spearman's rank correlation coefficients (Spearman 1904) and two-tailed P values were estimated using 'cor.test' function in R

ANOVA analysis

For multi-class clinical features (ordinal or nominal), one-way analysis of variance (Howell 2002) was applied to compare the log2-expression levels between different clinical classes using 'anova' function in R

Student's t-test analysis

For two-class clinical features, two-tailed Student's t test with unequal variance (Lehmann and Romano 2005) was applied to compare the log2-expression levels between the two clinical classes using 't.test' function in R

Q value calculation

For multiple hypothesis correction, Q value is the False Discovery Rate (FDR) analogue of the P value (Benjamini and Hochberg 1995), defined as the minimum FDR at which the test may be called significant. We used the 'Benjamini and Hochberg' method of 'p.adjust' function in R to convert P values into Q values.

Download Results

In addition to the links below, the full results of the analysis summarized in this report can also be downloaded programmatically using firehose_get, or interactively from either the Broad GDAC website or TCGA Data Coordination Center Portal.

References
[1] Andersen and Gill, Cox's regression model for counting processes, a large sample study, Annals of Statistics 10(4):1100-1120 (1982)
[2] Spearman, C, The proof and measurement of association between two things, Amer. J. Psychol 15:72-101 (1904)
[3] Howell, D, Statistical Methods for Psychology. (5th ed.), Duxbury Press:324-5 (2002)
[4] Lehmann and Romano, Testing Statistical Hypotheses (3E ed.), New York: Springer. ISBN 0387988645 (2005)
[5] Benjamini and Hochberg, Controlling the false discovery rate: a practical and powerful approach to multiple testing, Journal of the Royal Statistical Society Series B 59:289-300 (1995)
Copyright © 2014 Broad Institute TCGA GDAC as part of the TCGA Research Network. All rights reserved.
Made with Nozzle