| Wiki Markup |
|---|
h5. Analysis of batch vs clinical traits
Number of clinical traits: 84
Number of batches based on tumor DNA methylation data (samples retrieved according to this pattern: "TCGA-..-....-0..-..D-....-05"): 24
Correlation between center and batches ('two'=center (second field in the patient barcode)):
{code:collapse=true}> table(batchID,two)
two
batchID A1 A2 A7 A8 AC AN AO AQ AR B6 BH C8 D8 E2 E9 EW GI GM HN
A00Y 0 3 7 66 0 12 2 0 0 0 4 0 0 0 0 0 0 0 0
A032 0 22 0 14 0 19 10 1 0 16 10 0 0 0 0 0 0 0 0
A058 0 2 1 3 0 0 11 0 0 4 26 0 0 0 0 0 0 0 0
A088 7 14 0 1 0 0 1 0 8 9 7 0 0 0 0 0 0 0 0
A10A 0 12 0 0 0 9 0 0 5 9 4 0 0 0 0 0 0 0 0
A10N 0 1 0 0 0 1 12 2 0 1 1 0 0 9 0 0 0 0 0
A10P 7 16 1 4 0 0 12 0 8 13 32 0 0 0 0 0 0 0 0
A112 1 11 1 0 0 0 5 0 3 5 18 20 9 15 0 0 0 0 0
A12E 0 0 0 0 0 0 0 0 1 0 20 2 0 21 0 0 0 0 0
A12R 0 0 3 0 0 0 0 0 15 0 14 0 0 11 0 0 0 0 0
A138 0 0 0 0 0 0 0 0 1 0 7 6 0 2 0 0 0 0 0
A13K 0 7 1 0 0 0 5 2 0 3 18 4 19 6 0 8 0 0 0
A145 6 0 0 0 0 0 1 0 0 0 0 0 0 10 4 19 0 0 0
A148 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
A14H 0 1 0 0 0 0 0 0 0 0 1 0 7 5 10 1 0 0 0
A14N 0 0 0 0 0 0 0 1 0 1 2 0 20 1 6 1 0 0 0
A161 0 0 0 0 2 0 0 0 0 1 3 0 3 2 17 0 0 0 0
A16A 0 4 6 0 1 0 0 0 22 0 1 3 0 0 9 0 0 0 0
A16G 0 3 0 0 0 0 0 0 0 0 1 8 13 0 3 0 1 0 0
A17F 0 0 0 0 4 0 0 0 0 0 0 0 1 0 0 3 0 0 0
A17Z 0 0 0 0 2 0 0 0 4 0 0 0 0 0 1 0 0 6 0
A18O 0 0 0 0 1 0 0 0 5 1 1 0 0 0 1 0 0 7 1
A19F 0 2 0 1 0 0 0 0 0 1 0 0 0 1 0 0 0 0 0
A19Z 0 0 0 0 5 0 0 0 1 0 1 0 0 1 0 0 0 0 0
{code}
Significant batch-clinical traits correlations (the entire list can be found [here|^BatchClinicalInfoCorrelationsBRCA.csv]):
{csv}
"BRCA_clinical_traits","DataType","NumberOfNAs","Test","Pvalue"
"tissue_prospective_collection_indicator","factor",35,"Pearson's Chi-squared test",4.47E-62
"tissue_retrospective_collection_indicator","factor",35,"Pearson's Chi-squared test",4.47E-62
"year_of_initial_pathologic_diagnosis","integer",34,"Kruskal-Wallis rank sum test",3.15E-32
"breast_carcinoma_first_surgical_procedure_name","factor",54,"Pearson's Chi-squared test",5.45E-32
"days_to_last_followup","integer",73,"Kruskal-Wallis rank sum test",3.07E-31
"days_to_form_completion","integer",34,"Kruskal-Wallis rank sum test",5.70E-31
"first_pathologic_diagnosis_biospecimen_acquisition_method_type","factor",123,"Pearson's Chi-squared test",3.39E-28
"breast_tumor_clinical_m_stage","factor",35,"Pearson's Chi-squared test",1.06E-22
"axillary_lymph_node_stage_method_type","factor",223,"Pearson's Chi-squared test",9.33E-19
"breast_tumor_pathologic_n_stage","factor",34,"Pearson's Chi-squared test",2.19E-17
"lab_proc_her2_neu_immunohistochemistry_receptor_status","factor",41,"Pearson's Chi-squared test",6.22E-16
"breast_carcinoma_estrogen_receptor_status","factor",34,"Pearson's Chi-squared test",1.85E-13
"breast_carcinoma_progesterone_receptor_status","factor",34,"Pearson's Chi-squared test",8.87E-13
"vital_status","factor",34,"Pearson's Chi-squared test",2.38E-09
"anatomic_site_location_descriptor","factor",119,"Pearson's Chi-squared test",1.03E-07
"age_at_initial_pathologic_diagnosis","integer",34,"Kruskal-Wallis rank sum test",5.87E-06
"days_to_birth","integer",34,"Kruskal-Wallis rank sum test",6.68E-06
"lab_procedure_her2_neu_in_situ_hybrid_outcome_type","factor",194,"Pearson's Chi-squared test",3.18E-05
"person_menopause_status","factor",161,"Pearson's Chi-squared test",5.70E-05
"breast_tumor_pathologic_grouping_stage","factor",40,"Pearson's Chi-squared test",7.40E-05
"her2_immunohistochemistry_level_result","factor",351,"Pearson's Chi-squared test",1.72E-04
"breast_tumor_pathologic_t_stage","factor",34,"Pearson's Chi-squared test",2.82E-04
"pos_finding_lymph_node_hematoxylin_and_eosin_staining_microscopy_count","integer",177,"Kruskal-Wallis rank sum test",6.49E-04
"cytokeratin_immunohistochemistry_staining_method_micrometastasis_indicator","factor",324,"Pearson's Chi-squared test",8.61E-04
"person_neoplasm_cancer_status","factor",284,"Pearson's Chi-squared test",7.95E-03
"breast_cancer_optical_measurement_histologic_type","factor",34,"Pearson's Chi-squared test",1.47E-02
"disease_surgical_margin_status","factor",82,"Pearson's Chi-squared test",3.70E-02
{csv}
h5. Correlation with survival
Relevant clinical traits: days to the last follow-up (27), vital status (83), days to death (24), days to last know alive (28), summaries:
{code:collapse=true}> summary(clinical[,27]) # days to the last follow up
Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
0.0 140.0 457.0 815.8 1194.0 6795.0 73.0
> table(clinical[,83]) #vital status
DECEASED LIVING
93 725
> summary(clinical[,24]) # days to death
Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
157 811 1563 1744 2520 4456 759
> summary(clinical[,28]) # days to last known alive
Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
0.0 293.5 607.5 1068.0 1442.0 6795.0 508.0{code}
It seems that similarly to the colon cancer combined datasets days to last known alive is similar to the days to the last follow-up, however days to the last follow up contains more information (fewer NAs), use it for construction of the survival object. No patients missed information for both days to the last follow up and days to death. The survival object was created in the same way as for the analyses of other TCGA cancer datasets. Info is available ([here|METHYLATION:Colon cancer] and [here|METHYLATION:AML - acute myeloid leukemia])
Kaplan Meier curve and survival plots break down by batch:
!BRCA_KaplanMeierCurve.png|thumbnail! !BRCA_SurvivalVsBatch.png|thumbnail!
Here is the summary of the survival vs batch:
{code:collapse=true}
> summary(coxph(surv~methM[,2]))
Call:
coxph(formula = surv ~ methM[, 2])
n= 818, number of events= 93
(34 observations deleted due to missingness)
coef exp(coef) se(coef) z Pr(>|z|)
methM[, 2]A032 -6.961e-01 4.985e-01 5.148e-01 -1.352 0.1763
methM[, 2]A058 -2.321e+00 9.819e-02 1.084e+00 -2.140 0.0323 *
methM[, 2]A088 -5.512e-01 5.763e-01 5.413e-01 -1.018 0.3086
methM[, 2]A10A -2.238e-01 7.995e-01 5.565e-01 -0.402 0.6876
methM[, 2]A10N -1.455e+00 2.334e-01 8.217e-01 -1.771 0.0766 .
methM[, 2]A112 -9.643e-01 3.812e-01 5.847e-01 -1.649 0.0991 .
methM[, 2]A12E 9.108e-01 2.486e+00 5.015e-01 1.816 0.0694 .
methM[, 2]A12R -1.794e+00 1.663e-01 1.082e+00 -1.657 0.0975 .
methM[, 2]A138 8.921e-01 2.440e+00 5.487e-01 1.626 0.1040
methM[, 2]A13K 2.542e-01 1.289e+00 4.825e-01 0.527 0.5983
methM[, 2]A145 -9.748e-01 3.773e-01 1.081e+00 -0.902 0.3673
methM[, 2]A14H 3.164e-01 1.372e+00 8.215e-01 0.385 0.7002
methM[, 2]A14N 9.591e-01 2.609e+00 1.089e+00 0.880 0.3786
methM[, 2]A161 -1.382e-01 8.709e-01 7.183e-01 -0.192 0.8475
methM[, 2]A16A -1.347e+00 2.600e-01 8.206e-01 -1.642 0.1007
methM[, 2]A16G -1.714e+01 3.615e-08 5.030e+03 -0.003 0.9973
methM[, 2]A17F -1.731e+01 3.039e-08 1.338e+04 -0.001 0.9990
methM[, 2]A17Z -1.725e+01 3.216e-08 4.025e+03 -0.004 0.9966
methM[, 2]A18O -1.434e+00 2.383e-01 1.088e+00 -1.318 0.1875
methM[, 2]A19Z NA NA 0.000e+00 NA NA
---
Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
exp(coef) exp(-coef) lower .95 upper .95
methM[, 2]A032 4.985e-01 2.006e+00 0.18176 1.3672
methM[, 2]A058 9.819e-02 1.018e+01 0.01172 0.8223
methM[, 2]A088 5.763e-01 1.735e+00 0.19946 1.6649
methM[, 2]A10A 7.995e-01 1.251e+00 0.26858 2.3797
methM[, 2]A10N 2.334e-01 4.284e+00 0.04664 1.1685
methM[, 2]A112 3.812e-01 2.623e+00 0.12121 1.1991
methM[, 2]A12E 2.486e+00 4.022e-01 0.93037 6.6440
methM[, 2]A12R 1.663e-01 6.012e+00 0.01993 1.3879
methM[, 2]A138 2.440e+00 4.098e-01 0.83256 7.1522
methM[, 2]A13K 1.289e+00 7.755e-01 0.50082 3.3197
methM[, 2]A145 3.773e-01 2.651e+00 0.04531 3.1409
methM[, 2]A14H 1.372e+00 7.288e-01 0.27424 6.8653
methM[, 2]A14N 2.609e+00 3.832e-01 0.30851 22.0693
methM[, 2]A161 8.709e-01 1.148e+00 0.21309 3.5598
methM[, 2]A16A 2.600e-01 3.846e+00 0.05206 1.2985
methM[, 2]A16G 3.615e-08 2.766e+07 0.00000 Inf
methM[, 2]A17F 3.039e-08 3.290e+07 0.00000 Inf
methM[, 2]A17Z 3.216e-08 3.109e+07 0.00000 Inf
methM[, 2]A18O 2.383e-01 4.196e+00 0.02825 2.0108
methM[, 2]A19Z NA NA NA NA
Rsquare= 0.069 (max possible= 0.67 )
Likelihood ratio test= 58.61 on 19 df, p=6.4e-06
Wald test = 50 on 19 df, p=0.0001311
Score (logrank) test = 69.46 on 19 df, p=1.129e-07
Warning messages:
1: In fitter(X, Y, strats, offset, init, control, weights = weights, :
Loglik converged before variable 16,17,18 ; beta may be infinite.
2: In coxph(surv ~ methM[, 2]) :
X matrix deemed to be singular; variable 20{code}
_It seems that there are a lot of errors, I wonder why. I also don't understand where those observations come from that are deleted due to missingness. Need to ask someone to help clarify this output_.
h5. DNA methylation data
December 21st, 2011: 27k and 450k arrays are available. Downloaded Level 1 450k data. It seems that they started splitting green and red probes into 2 separate files and they also provide now the Illumina's idat files which are the bead level data (not tab delimited files). I need to find a way to process them, it seems that Bioconductor beadarray package can be used to read these files and do some bead level normalization (summarization too?). The Level2 data contains already summarized and normalized data (tab delimited files with CpG ID, value for methylated and value for unmethylated probes), however it is available only for 91 patients. Also tried to download 27k arrays available for breast cancer, however the data is available for \~26 patients (they stopped running those arrays?). I guess I need to figure out how to process Level 1 data. |
Page Comparison
General
Content
Integrations
App links