Correlation between aggregated molecular cancer subtypes and selected clinical features

Liver Hepatocellular Carcinoma

14 July 2016 | awg_lihc__2016_07_14

Maintainer Information

Citation Information

Maintained by TCGA GDAC Team (Broad Institute/MD Anderson Cancer Center/Harvard Medical School)

Cite as Broad Institute TCGA Genome Data Analysis Center (2016): Correlation between aggregated molecular cancer subtypes and selected clinical features. Broad Institute of MIT and Harvard. doi:10.7908/C1P26XNX

Overview

Introduction

This pipeline computes the correlation between cancer subtypes identified by different molecular patterns and selected clinical features.

Summary

Testing the association between subtypes identified by 10 different clustering approaches and 12 clinical features across 377 patients, 40 significant findings detected with P value < 0.05 and Q value < 0.25.

3 subtypes identified in current cancer cohort by 'Copy Number Ratio CNMF subtypes'. These subtypes correlate to 'RACE'.
4 subtypes identified in current cancer cohort by 'METHLYATION CNMF'. These subtypes correlate to 'YEARS_TO_BIRTH', 'GENDER', 'HISTOLOGICAL_TYPE', and 'RACE'.
CNMF clustering analysis on RPPA data identified 5 subtypes that correlate to 'YEARS_TO_BIRTH', 'PATHOLOGIC_STAGE', 'PATHOLOGY_T_STAGE', and 'RACE'.
Consensus hierarchical clustering analysis on RPPA data identified 5 subtypes that correlate to 'Time to Death', 'YEARS_TO_BIRTH', 'PATHOLOGIC_STAGE', 'PATHOLOGY_T_STAGE', 'GENDER', and 'RACE'.
CNMF clustering analysis on sequencing-based mRNA expression data identified 5 subtypes that correlate to 'YEARS_TO_BIRTH', 'PATHOLOGY_T_STAGE', 'GENDER', 'HISTOLOGICAL_TYPE', and 'RACE'.
Consensus hierarchical clustering analysis on sequencing-based mRNA expression data identified 4 subtypes that correlate to 'YEARS_TO_BIRTH', 'PATHOLOGY_T_STAGE', 'GENDER', and 'RACE'.
7 subtypes identified in current cancer cohort by 'MIRSEQ CNMF'. These subtypes correlate to 'YEARS_TO_BIRTH', 'GENDER', 'HISTOLOGICAL_TYPE', 'RACE', and 'ETHNICITY'.
6 subtypes identified in current cancer cohort by 'MIRSEQ CHIERARCHICAL'. These subtypes correlate to 'Time to Death', 'YEARS_TO_BIRTH', 'GENDER', and 'RACE'.
7 subtypes identified in current cancer cohort by 'MIRseq Mature CNMF subtypes'. These subtypes correlate to 'YEARS_TO_BIRTH', 'PATHOLOGY_T_STAGE', 'GENDER', 'HISTOLOGICAL_TYPE', and 'RACE'.
6 subtypes identified in current cancer cohort by 'MIRseq Mature cHierClus subtypes'. These subtypes correlate to 'YEARS_TO_BIRTH' and 'GENDER'.

Results

Overview of the results

Table 1. Get Full Table Overview of the association between subtypes identified by 10 different clustering approaches and 12 clinical features. Shown in the table are P values (Q values). Thresholded by P value < 0.05 and Q value < 0.25, 40 significant findings detected.


Clinical Features	Statistical Tests	Copy Number Ratio CNMF subtypes	METHLYATION CNMF	RPPA CNMF subtypes	RPPA cHierClus subtypes	RNAseq CNMF subtypes	RNAseq cHierClus subtypes	MIRSEQ CNMF	MIRSEQ CHIERARCHICAL	MIRseq Mature CNMF subtypes	MIRseq Mature cHierClus subtypes
Time to Death	logrank test	0.301 (0.532)	0.311 (0.534)	0.563 (0.709)	0.0429 (0.134)	0.197 (0.408)	0.397 (0.574)	0.268 (0.495)	0.0237 (0.0863)	0.184 (0.387)	0.09 (0.245)
YEARS TO BIRTH	Kruskal-Wallis (anova)	0.911 (0.926)	9.21e-07 (5.53e-05)	0.00207 (0.0146)	0.0495 (0.149)	0.000866 (0.00866)	1.44e-05 (4e-04)	0.000325 (0.00391)	8.53e-07 (5.53e-05)	3.89e-06 (0.000156)	5.99e-05 (0.000899)
PATHOLOGIC STAGE	Fisher's exact test	0.11 (0.269)	0.306 (0.532)	5e-05 (0.000857)	0.00147 (0.0126)	0.11 (0.269)	0.129 (0.304)	0.338 (0.556)	0.689 (0.78)	0.0598 (0.175)	0.113 (0.271)
PATHOLOGY T STAGE	Fisher's exact test	0.0678 (0.194)	0.514 (0.67)	0.00253 (0.0169)	0.00204 (0.0146)	0.0112 (0.0464)	0.0102 (0.0438)	0.431 (0.602)	0.687 (0.78)	0.0258 (0.0883)	0.366 (0.574)
PATHOLOGY N STAGE	Fisher's exact test	0.361 (0.574)	0.55 (0.702)	0.268 (0.495)	0.724 (0.805)	0.416 (0.594)	0.169 (0.361)	0.802 (0.837)	0.573 (0.709)	0.509 (0.67)	0.384 (0.574)
PATHOLOGY M STAGE	Fisher's exact test	0.29 (0.519)	0.387 (0.574)	0.285 (0.519)	0.641 (0.761)	0.766 (0.817)	0.679 (0.78)	0.0812 (0.226)	0.338 (0.556)	0.631 (0.757)	0.751 (0.817)
GENDER	Fisher's exact test	0.68 (0.78)	0.00296 (0.0178)	0.253 (0.49)	0.0434 (0.134)	2e-05 (4e-04)	2e-05 (4e-04)	0.0125 (0.0501)	0.0094 (0.0418)	0.0295 (0.0983)	0.0388 (0.126)
RADIATION THERAPY	Fisher's exact test	0.915 (0.926)	0.44 (0.608)	0.492 (0.658)	0.856 (0.885)	0.142 (0.324)	0.0965 (0.252)	0.766 (0.817)	0.202 (0.411)	0.72 (0.805)	0.15 (0.332)
HISTOLOGICAL TYPE	Fisher's exact test	0.101 (0.257)	0.00525 (0.0269)	0.494 (0.658)	0.092 (0.245)	0.00403 (0.022)	0.143 (0.324)	0.004 (0.022)	0.571 (0.709)	0.00539 (0.0269)	0.206 (0.411)
RESIDUAL TUMOR	Fisher's exact test	0.742 (0.817)	0.422 (0.596)	0.53 (0.684)	0.605 (0.734)	0.328 (0.554)	0.361 (0.574)	0.688 (0.78)	0.769 (0.817)	0.395 (0.574)	0.588 (0.72)
RACE	Fisher's exact test	0.0248 (0.0875)	0.00626 (0.0289)	0.00098 (0.00905)	0.00274 (0.0173)	0.00597 (0.0287)	0.00025 (0.00333)	0.00188 (0.0146)	0.0148 (0.0572)	0.00046 (0.00502)	0.387 (0.574)
ETHNICITY	Fisher's exact test	0.217 (0.426)	0.491 (0.658)	0.155 (0.339)	0.796 (0.837)	0.389 (0.574)	0.258 (0.492)	0.0198 (0.0743)	0.385 (0.574)	0.973 (0.973)	0.919 (0.926)

Clustering Approach #1: 'Copy Number Ratio CNMF subtypes'

Table S1. Description of clustering approach #1: 'Copy Number Ratio CNMF subtypes'

Cluster Labels	1	2	3
Number of samples	101	131	138

'Copy Number Ratio CNMF subtypes' versus 'Time to Death'

P value = 0.301 (logrank test), Q value = 0.53

Table S2. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	364	129	0.0 - 120.8 (19.8)
subtype1	99	29	0.3 - 108.8 (19.8)
subtype2	129	50	0.0 - 120.8 (19.3)
subtype3	136	50	0.2 - 114.3 (20.5)

Figure S1. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'Copy Number Ratio CNMF subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.911 (Kruskal-Wallis (anova)), Q value = 0.93

Table S3. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	366	59.4 (13.2)
subtype1	100	58.3 (14.7)
subtype2	130	60.2 (12.3)
subtype3	136	59.6 (12.8)

Figure S2. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'Copy Number Ratio CNMF subtypes' versus 'PATHOLOGIC_STAGE'

P value = 0.11 (Fisher's exact test), Q value = 0.27

Table S4. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV	STAGE IVA	STAGE IVB
ALL	173	87	3	62	9	7	2	1	2
subtype1	53	16	1	20	1	3	0	1	1
subtype2	52	37	1	21	5	4	0	0	0
subtype3	68	34	1	21	3	0	2	0	1

Figure S3. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'Copy Number Ratio CNMF subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.0678 (Fisher's exact test), Q value = 0.19

Table S5. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	183	95	77	12
subtype1	56	16	22	4
subtype2	55	43	27	6
subtype3	72	36	28	2

Figure S4. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'Copy Number Ratio CNMF subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.361 (Fisher's exact test), Q value = 0.57

Table S6. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	254	3
subtype1	68	1
subtype2	84	2
subtype3	102	0

Figure S5. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'Copy Number Ratio CNMF subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.29 (Fisher's exact test), Q value = 0.52

Table S7. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	266	4
subtype1	64	1
subtype2	96	0
subtype3	106	3

Figure S6. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'Copy Number Ratio CNMF subtypes' versus 'GENDER'

P value = 0.68 (Fisher's exact test), Q value = 0.78

Table S8. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	118	252
subtype1	29	72
subtype2	45	86
subtype3	44	94

Figure S7. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #7: 'GENDER'

'Copy Number Ratio CNMF subtypes' versus 'RADIATION_THERAPY'

P value = 0.915 (Fisher's exact test), Q value = 0.93

Table S9. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	338	9
subtype1	92	3
subtype2	119	3
subtype3	127	3

Figure S8. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

'Copy Number Ratio CNMF subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 0.101 (Fisher's exact test), Q value = 0.26

Table S10. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

nPatients	FIBROLAMELLAR CARCINOMA	HEPATOCELLULAR CARCINOMA	HEPATOCHOLANGIOCARCINOMA (MIXED)
ALL	3	360	7
subtype1	3	97	1
subtype2	0	129	2
subtype3	0	134	4

Figure S9. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

'Copy Number Ratio CNMF subtypes' versus 'RESIDUAL_TUMOR'

P value = 0.742 (Fisher's exact test), Q value = 0.82

Table S11. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	324	17	1	21
subtype1	85	5	0	7
subtype2	115	5	0	9
subtype3	124	7	1	5

Figure S10. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

'Copy Number Ratio CNMF subtypes' versus 'RACE'

P value = 0.0248 (Fisher's exact test), Q value = 0.088

Table S12. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #11: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	2	158	17	183
subtype1	0	31	6	61
subtype2	1	58	8	62
subtype3	1	69	3	60

Figure S11. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #11: 'RACE'

'Copy Number Ratio CNMF subtypes' versus 'ETHNICITY'

P value = 0.217 (Fisher's exact test), Q value = 0.43

Table S13. Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	16	335
subtype1	3	92
subtype2	9	115
subtype3	4	128

Figure S12. Get High-res Image Clustering Approach #1: 'Copy Number Ratio CNMF subtypes' versus Clinical Feature #12: 'ETHNICITY'

Clustering Approach #2: 'METHLYATION CNMF'

Table S14. Description of clustering approach #2: 'METHLYATION CNMF'

Cluster Labels	1	2	3	4
Number of samples	61	165	104	47

'METHLYATION CNMF' versus 'Time to Death'

P value = 0.311 (logrank test), Q value = 0.53

Table S15. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	371	132	0.0 - 120.8 (19.8)
subtype1	59	25	0.0 - 90.7 (20.7)
subtype2	164	54	0.2 - 120.8 (20.2)
subtype3	102	39	0.2 - 114.3 (18.2)
subtype4	46	14	0.2 - 108.8 (21.0)

Figure S13. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #1: 'Time to Death'

'METHLYATION CNMF' versus 'YEARS_TO_BIRTH'

P value = 9.21e-07 (Kruskal-Wallis (anova)), Q value = 5.5e-05

Table S16. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	373	59.3 (13.4)
subtype1	59	60.9 (12.2)
subtype2	164	63.1 (10.4)
subtype3	104	55.1 (14.6)
subtype4	46	53.0 (16.9)

Figure S14. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'METHLYATION CNMF' versus 'PATHOLOGIC_STAGE'

P value = 0.306 (Fisher's exact test), Q value = 0.53

Table S17. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV	STAGE IVA	STAGE IVB
ALL	175	87	3	65	9	9	2	1	2
subtype1	24	15	0	11	4	4	0	0	0
subtype2	81	39	2	26	3	3	0	0	1
subtype3	48	25	1	15	2	2	2	0	1
subtype4	22	8	0	13	0	0	0	1	0

Figure S15. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'METHLYATION CNMF' versus 'PATHOLOGY_T_STAGE'

P value = 0.514 (Fisher's exact test), Q value = 0.67

Table S18. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	185	95	81	13
subtype1	24	19	15	3
subtype2	86	41	31	6
subtype3	52	27	21	4
subtype4	23	8	14	0

Figure S16. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'METHLYATION CNMF' versus 'PATHOLOGY_N_STAGE'

P value = 0.55 (Fisher's exact test), Q value = 0.7

Table S19. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	257	4
subtype1	40	1
subtype2	111	1
subtype3	77	1
subtype4	29	1

Figure S17. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'METHLYATION CNMF' versus 'PATHOLOGY_M_STAGE'

P value = 0.387 (Fisher's exact test), Q value = 0.57

Table S20. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	272	4
subtype1	43	0
subtype2	118	1
subtype3	80	3
subtype4	31	0

Figure S18. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'METHLYATION CNMF' versus 'GENDER'

P value = 0.00296 (Fisher's exact test), Q value = 0.018

Table S21. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	122	255
subtype1	22	39
subtype2	38	127
subtype3	46	58
subtype4	16	31

Figure S19. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #7: 'GENDER'

'METHLYATION CNMF' versus 'RADIATION_THERAPY'

P value = 0.44 (Fisher's exact test), Q value = 0.61

Table S22. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	345	9
subtype1	58	0
subtype2	154	4
subtype3	92	3
subtype4	41	2

Figure S20. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #8: 'RADIATION_THERAPY'

'METHLYATION CNMF' versus 'HISTOLOGICAL_TYPE'

P value = 0.00525 (Fisher's exact test), Q value = 0.027

Table S23. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

nPatients	FIBROLAMELLAR CARCINOMA	HEPATOCELLULAR CARCINOMA	HEPATOCHOLANGIOCARCINOMA (MIXED)
ALL	3	367	7
subtype1	0	59	2
subtype2	0	164	1
subtype3	0	101	3
subtype4	3	43	1

Figure S21. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

'METHLYATION CNMF' versus 'RESIDUAL_TUMOR'

P value = 0.422 (Fisher's exact test), Q value = 0.6

Table S24. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	330	17	1	22
subtype1	48	5	0	5
subtype2	147	7	0	8
subtype3	91	5	1	7
subtype4	44	0	0	2

Figure S22. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

'METHLYATION CNMF' versus 'RACE'

P value = 0.00626 (Fisher's exact test), Q value = 0.029

Table S25. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #11: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	2	161	17	187
subtype1	0	20	6	33
subtype2	1	66	8	86
subtype3	1	60	1	40
subtype4	0	15	2	28

Figure S23. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #11: 'RACE'

'METHLYATION CNMF' versus 'ETHNICITY'

P value = 0.491 (Fisher's exact test), Q value = 0.66

Table S26. Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	18	340
subtype1	4	55
subtype2	5	149
subtype3	7	95
subtype4	2	41

Figure S24. Get High-res Image Clustering Approach #2: 'METHLYATION CNMF' versus Clinical Feature #12: 'ETHNICITY'

Clustering Approach #3: 'RPPA CNMF subtypes'

Table S27. Description of clustering approach #3: 'RPPA CNMF subtypes'

Cluster Labels	1	2	3	4	5	6
Number of samples	41	75	19	23	24	2

'RPPA CNMF subtypes' versus 'Time to Death'

P value = 0.563 (logrank test), Q value = 0.71

Table S28. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	180	93	0.2 - 114.3 (17.6)
subtype1	41	24	0.2 - 113.0 (13.6)
subtype2	75	35	0.3 - 108.8 (20.0)
subtype3	19	12	0.3 - 114.3 (26.9)
subtype4	22	9	0.5 - 66.3 (13.4)
subtype5	23	13	0.3 - 58.5 (14.6)

Figure S25. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'RPPA CNMF subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.00207 (Kruskal-Wallis (anova)), Q value = 0.015

Table S29. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	180	60.2 (14.8)
subtype1	39	61.9 (11.8)
subtype2	75	64.5 (12.8)
subtype3	19	56.5 (17.0)
subtype4	23	53.8 (15.7)
subtype5	24	53.3 (18.1)

Figure S26. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'RPPA CNMF subtypes' versus 'PATHOLOGIC_STAGE'

P value = 5e-05 (Fisher's exact test), Q value = 0.00086

Table S30. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV	STAGE IVB
ALL	70	38	3	44	6	6	2	1
subtype1	9	10	0	11	4	4	0	0
subtype2	43	10	2	13	0	0	1	0
subtype3	6	6	0	4	0	0	0	1
subtype4	9	3	0	9	0	0	1	0
subtype5	3	9	1	7	2	2	0	0

Figure S27. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'RPPA CNMF subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.00253 (Fisher's exact test), Q value = 0.017

Table S31. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	73	43	56	9
subtype1	10	12	15	4
subtype2	44	11	17	2
subtype3	7	6	5	1
subtype4	9	5	9	0
subtype5	3	9	10	2

Figure S28. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'RPPA CNMF subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.268 (Fisher's exact test), Q value = 0.49

Table S32. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	110	3
subtype1	25	2
subtype2	37	0
subtype3	17	0
subtype4	15	0
subtype5	16	1

Figure S29. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'RPPA CNMF subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.285 (Fisher's exact test), Q value = 0.52

Table S33. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	130	3
subtype1	35	0
subtype2	46	1
subtype3	16	1
subtype4	13	1
subtype5	20	0

Figure S30. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'RPPA CNMF subtypes' versus 'GENDER'

P value = 0.253 (Fisher's exact test), Q value = 0.49

Table S34. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	69	113
subtype1	16	25
subtype2	22	53
subtype3	9	10
subtype4	12	11
subtype5	10	14

Figure S31. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #7: 'GENDER'

'RPPA CNMF subtypes' versus 'RADIATION_THERAPY'

P value = 0.492 (Fisher's exact test), Q value = 0.66

Table S35. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	163	3
subtype1	38	1
subtype2	64	1
subtype3	16	1
subtype4	23	0
subtype5	22	0

Figure S32. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

'RPPA CNMF subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 0.494 (Fisher's exact test), Q value = 0.66

Table S36. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

nPatients	FIBROLAMELLAR CARCINOMA	HEPATOCELLULAR CARCINOMA	HEPATOCHOLANGIOCARCINOMA (MIXED)
ALL	2	178	2
subtype1	1	38	2
subtype2	1	74	0
subtype3	0	19	0
subtype4	0	23	0
subtype5	0	24	0

Figure S33. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

'RPPA CNMF subtypes' versus 'RESIDUAL_TUMOR'

P value = 0.53 (Fisher's exact test), Q value = 0.68

Table S37. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	147	11	1	16
subtype1	30	5	0	3
subtype2	61	5	1	4
subtype3	17	0	0	2
subtype4	20	0	0	3
subtype5	19	1	0	4

Figure S34. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

'RPPA CNMF subtypes' versus 'RACE'

P value = 0.00098 (Fisher's exact test), Q value = 0.009

Table S38. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #11: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	52	12	111
subtype1	14	0	24
subtype2	9	7	56
subtype3	9	1	9
subtype4	10	1	11
subtype5	10	3	11

Figure S35. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #11: 'RACE'

'RPPA CNMF subtypes' versus 'ETHNICITY'

P value = 0.155 (Fisher's exact test), Q value = 0.34

Table S39. Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	7	161
subtype1	2	34
subtype2	2	67
subtype3	0	19
subtype4	3	18
subtype5	0	23

Figure S36. Get High-res Image Clustering Approach #3: 'RPPA CNMF subtypes' versus Clinical Feature #12: 'ETHNICITY'

Clustering Approach #4: 'RPPA cHierClus subtypes'

Table S40. Description of clustering approach #4: 'RPPA cHierClus subtypes'

Cluster Labels	1	2	3	4	5
Number of samples	41	47	38	19	39

'RPPA cHierClus subtypes' versus 'Time to Death'

P value = 0.0429 (logrank test), Q value = 0.13

Table S41. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	182	94	0.2 - 114.3 (17.8)
subtype1	40	19	0.3 - 114.3 (19.1)
subtype2	47	21	0.3 - 108.8 (19.8)
subtype3	37	22	0.2 - 81.7 (12.3)
subtype4	19	13	0.4 - 60.9 (13.6)
subtype5	39	19	0.3 - 90.7 (21.7)

Figure S37. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RPPA cHierClus subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.0495 (Kruskal-Wallis (anova)), Q value = 0.15

Table S42. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	182	60.4 (14.9)
subtype1	41	56.2 (18.3)
subtype2	47	63.2 (12.6)
subtype3	36	60.9 (10.3)
subtype4	19	53.8 (17.7)
subtype5	39	64.3 (13.8)

Figure S38. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'RPPA cHierClus subtypes' versus 'PATHOLOGIC_STAGE'

P value = 0.00147 (Fisher's exact test), Q value = 0.013

Table S43. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV	STAGE IVB
ALL	71	38	3	45	6	6	2	1
subtype1	11	10	1	10	2	4	1	1
subtype2	28	6	1	6	0	0	1	0
subtype3	7	12	0	11	3	1	0	0
subtype4	4	5	0	8	1	0	0	0
subtype5	21	5	1	10	0	1	0	0

Figure S39. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'RPPA cHierClus subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.00204 (Fisher's exact test), Q value = 0.015

Table S44. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	74	43	57	9
subtype1	11	12	15	3
subtype2	29	7	9	1
subtype3	8	14	14	2
subtype4	4	5	8	2
subtype5	22	5	11	1

Figure S40. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'RPPA cHierClus subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.724 (Fisher's exact test), Q value = 0.8

Table S45. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	112	3
subtype1	32	2
subtype2	20	0
subtype3	22	1
subtype4	12	0
subtype5	26	0

Figure S41. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'RPPA cHierClus subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.641 (Fisher's exact test), Q value = 0.76

Table S46. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	132	3
subtype1	32	2
subtype2	28	1
subtype3	32	0
subtype4	14	0
subtype5	26	0

Figure S42. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'RPPA cHierClus subtypes' versus 'GENDER'

P value = 0.0434 (Fisher's exact test), Q value = 0.13

Table S47. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	70	114
subtype1	24	17
subtype2	13	34
subtype3	13	25
subtype4	7	12
subtype5	13	26

Figure S43. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

'RPPA cHierClus subtypes' versus 'RADIATION_THERAPY'

P value = 0.856 (Fisher's exact test), Q value = 0.89

Table S48. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	165	3
subtype1	39	0
subtype2	40	1
subtype3	34	1
subtype4	18	0
subtype5	34	1

Figure S44. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

'RPPA cHierClus subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 0.092 (Fisher's exact test), Q value = 0.25

Table S49. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

nPatients	FIBROLAMELLAR CARCINOMA	HEPATOCELLULAR CARCINOMA	HEPATOCHOLANGIOCARCINOMA (MIXED)
ALL	2	180	2
subtype1	0	39	2
subtype2	0	47	0
subtype3	0	38	0
subtype4	0	19	0
subtype5	2	37	0

Figure S45. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

'RPPA cHierClus subtypes' versus 'RESIDUAL_TUMOR'

P value = 0.605 (Fisher's exact test), Q value = 0.73

Table S50. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	148	12	1	16
subtype1	34	2	0	4
subtype2	37	4	1	2
subtype3	27	4	0	6
subtype4	17	0	0	2
subtype5	33	2	0	2

Figure S46. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

'RPPA cHierClus subtypes' versus 'RACE'

P value = 0.00274 (Fisher's exact test), Q value = 0.017

Table S51. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #11: 'RACE'

nPatients	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	52	12	113
subtype1	15	1	25
subtype2	4	4	36
subtype3	17	1	18
subtype4	8	2	9
subtype5	8	4	25

Figure S47. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #11: 'RACE'

'RPPA cHierClus subtypes' versus 'ETHNICITY'

P value = 0.796 (Fisher's exact test), Q value = 0.84

Table S52. Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	8	162
subtype1	3	37
subtype2	1	42
subtype3	1	33
subtype4	1	17
subtype5	2	33

Figure S48. Get High-res Image Clustering Approach #4: 'RPPA cHierClus subtypes' versus Clinical Feature #12: 'ETHNICITY'

Clustering Approach #5: 'RNAseq CNMF subtypes'

Table S53. Description of clustering approach #5: 'RNAseq CNMF subtypes'

Cluster Labels	1	2	3	4	5
Number of samples	129	90	50	73	29

'RNAseq CNMF subtypes' versus 'Time to Death'

P value = 0.197 (logrank test), Q value = 0.41

Table S54. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	365	130	0.0 - 120.8 (19.6)
subtype1	123	54	0.2 - 114.3 (17.6)
subtype2	90	32	0.3 - 120.8 (21.6)
subtype3	50	13	0.0 - 82.6 (17.0)
subtype4	73	23	0.4 - 108.8 (18.3)
subtype5	29	8	1.5 - 80.8 (25.3)

Figure S49. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq CNMF subtypes' versus 'YEARS_TO_BIRTH'

P value = 0.000866 (Kruskal-Wallis (anova)), Q value = 0.0087

Table S55. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	367	59.3 (13.4)
subtype1	127	56.0 (13.7)
subtype2	88	62.5 (11.4)
subtype3	50	58.0 (14.3)
subtype4	73	60.7 (13.3)
subtype5	29	62.4 (14.1)

Figure S50. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'RNAseq CNMF subtypes' versus 'PATHOLOGIC_STAGE'

P value = 0.11 (Fisher's exact test), Q value = 0.27

Table S56. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV	STAGE IVA	STAGE IVB
ALL	171	86	3	65	8	9	2	1	2
subtype1	46	36	1	29	4	5	1	0	1
subtype2	51	18	1	9	2	1	0	0	1
subtype3	29	9	0	4	0	2	1	1	0
subtype4	30	19	1	17	2	1	0	0	0
subtype5	15	4	0	6	0	0	0	0	0

Figure S51. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'RNAseq CNMF subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.0112 (Fisher's exact test), Q value = 0.046

Table S57. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	181	94	80	13
subtype1	48	40	36	5
subtype2	53	21	11	5
subtype3	32	9	7	2
subtype4	31	20	20	1
subtype5	17	4	6	0

Figure S52. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'RNAseq CNMF subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.416 (Fisher's exact test), Q value = 0.59

Table S58. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	252	4
subtype1	92	3
subtype2	61	0
subtype3	29	1
subtype4	54	0
subtype5	16	0

Figure S53. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'RNAseq CNMF subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.766 (Fisher's exact test), Q value = 0.82

Table S59. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	266	4
subtype1	102	2
subtype2	61	1
subtype3	30	1
subtype4	54	0
subtype5	19	0

Figure S54. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'RNAseq CNMF subtypes' versus 'GENDER'

P value = 2e-05 (Fisher's exact test), Q value = 4e-04

Table S60. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	121	250
subtype1	60	69
subtype2	27	63
subtype3	13	37
subtype4	10	63
subtype5	11	18

Figure S55. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #7: 'GENDER'

'RNAseq CNMF subtypes' versus 'RADIATION_THERAPY'

P value = 0.142 (Fisher's exact test), Q value = 0.32

Table S61. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	339	9
subtype1	118	1
subtype2	84	2
subtype3	44	1
subtype4	66	5
subtype5	27	0

Figure S56. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

'RNAseq CNMF subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 0.00403 (Fisher's exact test), Q value = 0.022

Table S62. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

nPatients	FIBROLAMELLAR CARCINOMA	HEPATOCELLULAR CARCINOMA	HEPATOCHOLANGIOCARCINOMA (MIXED)
ALL	3	361	7
subtype1	0	124	5
subtype2	0	90	0
subtype3	3	45	2
subtype4	0	73	0
subtype5	0	29	0

Figure S57. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

'RNAseq CNMF subtypes' versus 'RESIDUAL_TUMOR'

P value = 0.328 (Fisher's exact test), Q value = 0.55

Table S63. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	324	17	1	22
subtype1	111	7	0	10
subtype2	78	4	0	7
subtype3	40	4	1	3
subtype4	68	2	0	1
subtype5	27	0	0	1

Figure S58. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

'RNAseq CNMF subtypes' versus 'RACE'

P value = 0.00597 (Fisher's exact test), Q value = 0.029

Table S64. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #11: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	2	158	17	184
subtype1	1	72	2	52
subtype2	0	31	6	48
subtype3	0	17	2	29
subtype4	1	32	5	35
subtype5	0	6	2	20

Figure S59. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #11: 'RACE'

'RNAseq CNMF subtypes' versus 'ETHNICITY'

P value = 0.389 (Fisher's exact test), Q value = 0.57

Table S65. Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	18	334
subtype1	5	119
subtype2	8	75
subtype3	2	44
subtype4	2	69
subtype5	1	27

Figure S60. Get High-res Image Clustering Approach #5: 'RNAseq CNMF subtypes' versus Clinical Feature #12: 'ETHNICITY'

Clustering Approach #6: 'RNAseq cHierClus subtypes'

Table S66. Description of clustering approach #6: 'RNAseq cHierClus subtypes'

Cluster Labels	1	2	3	4
Number of samples	155	147	62	7

'RNAseq cHierClus subtypes' versus 'Time to Death'

P value = 0.397 (logrank test), Q value = 0.57

Table S67. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	365	130	0.0 - 120.8 (19.6)
subtype1	149	61	0.2 - 114.3 (18.3)
subtype2	147	49	0.0 - 120.8 (20.9)
subtype3	62	17	0.4 - 108.8 (19.0)
subtype4	7	3	10.3 - 80.8 (24.9)

Figure S61. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'RNAseq cHierClus subtypes' versus 'YEARS_TO_BIRTH'

P value = 1.44e-05 (Kruskal-Wallis (anova)), Q value = 4e-04

Table S68. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	367	59.3 (13.4)
subtype1	153	55.1 (14.6)
subtype2	145	62.7 (11.1)
subtype3	62	61.4 (13.3)
subtype4	7	61.7 (8.4)

Figure S62. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'RNAseq cHierClus subtypes' versus 'PATHOLOGIC_STAGE'

P value = 0.129 (Fisher's exact test), Q value = 0.3

Table S69. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV	STAGE IVA	STAGE IVB
ALL	171	86	3	65	8	9	2	1	2
subtype1	58	40	1	34	4	5	2	1	1
subtype2	85	28	1	15	3	3	0	0	1
subtype3	24	17	1	15	1	1	0	0	0
subtype4	4	1	0	1	0	0	0	0	0

Figure S63. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'RNAseq cHierClus subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.0102 (Fisher's exact test), Q value = 0.044

Table S70. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	181	94	80	13
subtype1	63	44	43	5
subtype2	89	31	19	7
subtype3	25	18	17	1
subtype4	4	1	1	0

Figure S64. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'RNAseq cHierClus subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.169 (Fisher's exact test), Q value = 0.36

Table S71. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	252	4
subtype1	110	4
subtype2	92	0
subtype3	48	0
subtype4	2	0

Figure S65. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'RNAseq cHierClus subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.679 (Fisher's exact test), Q value = 0.78

Table S72. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	266	4
subtype1	118	3
subtype2	98	1
subtype3	48	0
subtype4	2	0

Figure S66. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'RNAseq cHierClus subtypes' versus 'GENDER'

P value = 2e-05 (Fisher's exact test), Q value = 4e-04

Table S73. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	121	250
subtype1	68	87
subtype2	41	106
subtype3	8	54
subtype4	4	3

Figure S67. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

'RNAseq cHierClus subtypes' versus 'RADIATION_THERAPY'

P value = 0.0965 (Fisher's exact test), Q value = 0.25

Table S74. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	339	9
subtype1	139	4
subtype2	137	1
subtype3	56	4
subtype4	7	0

Figure S68. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

'RNAseq cHierClus subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 0.143 (Fisher's exact test), Q value = 0.32

Table S75. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

nPatients	FIBROLAMELLAR CARCINOMA	HEPATOCELLULAR CARCINOMA	HEPATOCHOLANGIOCARCINOMA (MIXED)
ALL	3	361	7
subtype1	3	146	6
subtype2	0	146	1
subtype3	0	62	0
subtype4	0	7	0

Figure S69. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

'RNAseq cHierClus subtypes' versus 'RESIDUAL_TUMOR'

P value = 0.361 (Fisher's exact test), Q value = 0.57

Table S76. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	324	17	1	22
subtype1	135	8	1	10
subtype2	124	7	0	12
subtype3	58	2	0	0
subtype4	7	0	0	0

Figure S70. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

'RNAseq cHierClus subtypes' versus 'RACE'

P value = 0.00025 (Fisher's exact test), Q value = 0.0033

Table S77. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #11: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	2	158	17	184
subtype1	1	85	3	64
subtype2	0	46	9	84
subtype3	1	27	4	30
subtype4	0	0	1	6

Figure S71. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #11: 'RACE'

'RNAseq cHierClus subtypes' versus 'ETHNICITY'

P value = 0.258 (Fisher's exact test), Q value = 0.49

Table S78. Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	18	334
subtype1	6	142
subtype2	11	126
subtype3	1	59
subtype4	0	7

Figure S72. Get High-res Image Clustering Approach #6: 'RNAseq cHierClus subtypes' versus Clinical Feature #12: 'ETHNICITY'

Clustering Approach #7: 'MIRSEQ CNMF'

Table S79. Description of clustering approach #7: 'MIRSEQ CNMF'

Cluster Labels	1	2	3	4	5	6	7
Number of samples	45	61	104	45	64	40	13

'MIRSEQ CNMF' versus 'Time to Death'

P value = 0.268 (logrank test), Q value = 0.49

Table S80. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	366	128	0.0 - 120.8 (19.8)
subtype1	43	20	0.0 - 120.8 (13.6)
subtype2	61	24	0.4 - 113.0 (21.8)
subtype3	103	30	0.3 - 108.8 (17.6)
subtype4	45	14	0.5 - 90.3 (29.8)
subtype5	62	23	0.2 - 114.3 (18.2)
subtype6	39	14	0.4 - 76.2 (20.8)
subtype7	13	3	10.3 - 70.1 (22.6)

Figure S73. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #1: 'Time to Death'

'MIRSEQ CNMF' versus 'YEARS_TO_BIRTH'

P value = 0.000325 (Kruskal-Wallis (anova)), Q value = 0.0039

Table S81. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	368	59.2 (13.4)
subtype1	43	60.3 (11.8)
subtype2	60	63.8 (11.0)
subtype3	104	61.8 (11.5)
subtype4	45	55.2 (14.4)
subtype5	63	56.4 (14.6)
subtype6	40	52.8 (15.8)
subtype7	13	60.6 (13.7)

Figure S74. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'MIRSEQ CNMF' versus 'PATHOLOGIC_STAGE'

P value = 0.338 (Fisher's exact test), Q value = 0.56

Table S82. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV	STAGE IVA	STAGE IVB
ALL	172	86	3	64	9	9	2	1	2
subtype1	20	8	0	12	2	1	0	0	1
subtype2	29	14	1	5	3	1	0	0	0
subtype3	52	26	0	17	3	2	0	0	0
subtype4	25	8	0	5	0	0	1	1	1
subtype5	23	17	1	15	1	2	0	0	0
subtype6	15	11	1	8	0	2	1	0	0
subtype7	8	2	0	2	0	1	0	0	0

Figure S75. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'MIRSEQ CNMF' versus 'PATHOLOGY_T_STAGE'

P value = 0.431 (Fisher's exact test), Q value = 0.6

Table S83. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	182	94	80	13
subtype1	20	9	14	2
subtype2	31	16	10	4
subtype3	53	27	20	3
subtype4	29	8	6	1
subtype5	24	20	19	1
subtype6	18	12	9	1
subtype7	7	2	2	1

Figure S76. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'MIRSEQ CNMF' versus 'PATHOLOGY_N_STAGE'

P value = 0.802 (Fisher's exact test), Q value = 0.84

Table S84. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	254	4
subtype1	32	0
subtype2	35	0
subtype3	73	1
subtype4	29	1
subtype5	47	1
subtype6	30	1
subtype7	8	0

Figure S77. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'MIRSEQ CNMF' versus 'PATHOLOGY_M_STAGE'

P value = 0.0812 (Fisher's exact test), Q value = 0.23

Table S85. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	269	4
subtype1	28	1
subtype2	40	0
subtype3	77	0
subtype4	31	2
subtype5	52	0
subtype6	33	1
subtype7	8	0

Figure S78. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'MIRSEQ CNMF' versus 'GENDER'

P value = 0.0125 (Fisher's exact test), Q value = 0.05

Table S86. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	119	253
subtype1	16	29
subtype2	21	40
subtype3	20	84
subtype4	14	31
subtype5	30	34
subtype6	15	25
subtype7	3	10

Figure S79. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #7: 'GENDER'

'MIRSEQ CNMF' versus 'RADIATION_THERAPY'

P value = 0.766 (Fisher's exact test), Q value = 0.82

Table S87. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	340	9
subtype1	43	0
subtype2	59	1
subtype3	93	4
subtype4	39	1
subtype5	59	1
subtype6	35	2
subtype7	12	0

Figure S80. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #8: 'RADIATION_THERAPY'

'MIRSEQ CNMF' versus 'HISTOLOGICAL_TYPE'

P value = 0.004 (Fisher's exact test), Q value = 0.022

Table S88. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

nPatients	FIBROLAMELLAR CARCINOMA	HEPATOCELLULAR CARCINOMA	HEPATOCHOLANGIOCARCINOMA (MIXED)
ALL	3	362	7
subtype1	0	45	0
subtype2	0	61	0
subtype3	0	103	1
subtype4	3	42	0
subtype5	0	59	5
subtype6	0	39	1
subtype7	0	13	0

Figure S81. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

'MIRSEQ CNMF' versus 'RESIDUAL_TUMOR'

P value = 0.688 (Fisher's exact test), Q value = 0.78

Table S89. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	325	17	1	22
subtype1	35	3	0	4
subtype2	54	3	0	3
subtype3	92	5	0	4
subtype4	39	3	1	2
subtype5	57	3	0	4
subtype6	35	0	0	5
subtype7	13	0	0	0

Figure S82. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

'MIRSEQ CNMF' versus 'RACE'

P value = 0.00188 (Fisher's exact test), Q value = 0.015

Table S90. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #11: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	2	161	17	182
subtype1	0	19	6	20
subtype2	0	16	4	35
subtype3	0	52	2	50
subtype4	0	18	4	20
subtype5	1	32	0	30
subtype6	0	22	1	17
subtype7	1	2	0	10

Figure S83. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #11: 'RACE'

'MIRSEQ CNMF' versus 'ETHNICITY'

P value = 0.0198 (Fisher's exact test), Q value = 0.074

Table S91. Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	18	335
subtype1	2	42
subtype2	6	49
subtype3	1	101
subtype4	0	39
subtype5	5	56
subtype6	4	35
subtype7	0	13

Figure S84. Get High-res Image Clustering Approach #7: 'MIRSEQ CNMF' versus Clinical Feature #12: 'ETHNICITY'

Clustering Approach #8: 'MIRSEQ CHIERARCHICAL'

Table S92. Description of clustering approach #8: 'MIRSEQ CHIERARCHICAL'

Cluster Labels	1	2	3	4	5	6
Number of samples	41	88	54	124	57	8

'MIRSEQ CHIERARCHICAL' versus 'Time to Death'

P value = 0.0237 (logrank test), Q value = 0.086

Table S93. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	366	128	0.0 - 120.8 (19.8)
subtype1	39	18	0.0 - 120.8 (13.7)
subtype2	88	30	0.3 - 102.7 (23.0)
subtype3	53	25	0.3 - 113.0 (14.9)
subtype4	122	40	0.2 - 114.3 (20.9)
subtype5	56	13	0.3 - 107.1 (17.7)
subtype6	8	2	10.3 - 108.8 (25.1)

Figure S85. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #1: 'Time to Death'

'MIRSEQ CHIERARCHICAL' versus 'YEARS_TO_BIRTH'

P value = 8.53e-07 (Kruskal-Wallis (anova)), Q value = 5.5e-05

Table S94. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	368	59.2 (13.4)
subtype1	40	61.7 (11.4)
subtype2	86	63.7 (11.2)
subtype3	54	59.8 (11.9)
subtype4	123	53.3 (15.5)
subtype5	57	63.3 (9.5)
subtype6	8	56.8 (11.9)

Figure S86. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'MIRSEQ CHIERARCHICAL' versus 'PATHOLOGIC_STAGE'

P value = 0.689 (Fisher's exact test), Q value = 0.78

Table S95. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV	STAGE IVA	STAGE IVB
ALL	172	86	3	64	9	9	2	1	2
subtype1	19	6	0	11	2	1	0	0	1
subtype2	45	19	1	8	3	1	0	0	0
subtype3	24	13	0	11	0	2	0	0	0
subtype4	56	31	2	22	1	3	2	1	1
subtype5	25	15	0	11	3	1	0	0	0
subtype6	3	2	0	1	0	1	0	0	0

Figure S87. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'MIRSEQ CHIERARCHICAL' versus 'PATHOLOGY_T_STAGE'

P value = 0.687 (Fisher's exact test), Q value = 0.78

Table S96. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	182	94	80	13
subtype1	19	7	13	2
subtype2	49	22	13	4
subtype3	26	14	12	2
subtype4	59	34	28	2
subtype5	26	15	13	2
subtype6	3	2	1	1

Figure S88. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'MIRSEQ CHIERARCHICAL' versus 'PATHOLOGY_N_STAGE'

P value = 0.573 (Fisher's exact test), Q value = 0.71

Table S97. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	254	4
subtype1	29	0
subtype2	50	0
subtype3	37	1
subtype4	89	3
subtype5	45	0
subtype6	4	0

Figure S89. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'MIRSEQ CHIERARCHICAL' versus 'PATHOLOGY_M_STAGE'

P value = 0.338 (Fisher's exact test), Q value = 0.56

Table S98. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	269	4
subtype1	25	1
subtype2	58	0
subtype3	41	0
subtype4	96	3
subtype5	46	0
subtype6	3	0

Figure S90. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'MIRSEQ CHIERARCHICAL' versus 'GENDER'

P value = 0.0094 (Fisher's exact test), Q value = 0.042

Table S99. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	119	253
subtype1	14	27
subtype2	24	64
subtype3	22	32
subtype4	48	76
subtype5	8	49
subtype6	3	5

Figure S91. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #7: 'GENDER'

'MIRSEQ CHIERARCHICAL' versus 'RADIATION_THERAPY'

P value = 0.202 (Fisher's exact test), Q value = 0.41

Table S100. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	340	9
subtype1	39	0
subtype2	83	1
subtype3	49	0
subtype4	111	4
subtype5	50	4
subtype6	8	0

Figure S92. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #8: 'RADIATION_THERAPY'

'MIRSEQ CHIERARCHICAL' versus 'HISTOLOGICAL_TYPE'

P value = 0.571 (Fisher's exact test), Q value = 0.71

Table S101. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

nPatients	FIBROLAMELLAR CARCINOMA	HEPATOCELLULAR CARCINOMA	HEPATOCHOLANGIOCARCINOMA (MIXED)
ALL	3	362	7
subtype1	0	41	0
subtype2	0	87	1
subtype3	0	52	2
subtype4	3	117	4
subtype5	0	57	0
subtype6	0	8	0

Figure S93. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

'MIRSEQ CHIERARCHICAL' versus 'RESIDUAL_TUMOR'

P value = 0.769 (Fisher's exact test), Q value = 0.82

Table S102. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	325	17	1	22
subtype1	32	3	0	3
subtype2	75	4	0	6
subtype3	48	4	0	2
subtype4	110	4	1	9
subtype5	53	1	0	2
subtype6	7	1	0	0

Figure S94. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

'MIRSEQ CHIERARCHICAL' versus 'RACE'

P value = 0.0148 (Fisher's exact test), Q value = 0.057

Table S103. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #11: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	2	161	17	182
subtype1	0	16	5	20
subtype2	0	28	5	50
subtype3	0	29	1	23
subtype4	1	60	3	56
subtype5	1	28	2	26
subtype6	0	0	1	7

Figure S95. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #11: 'RACE'

'MIRSEQ CHIERARCHICAL' versus 'ETHNICITY'

P value = 0.385 (Fisher's exact test), Q value = 0.57

Table S104. Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	18	335
subtype1	2	38
subtype2	4	77
subtype3	3	49
subtype4	9	109
subtype5	0	54
subtype6	0	8

Figure S96. Get High-res Image Clustering Approach #8: 'MIRSEQ CHIERARCHICAL' versus Clinical Feature #12: 'ETHNICITY'

Clustering Approach #9: 'MIRseq Mature CNMF subtypes'

Table S105. Description of clustering approach #9: 'MIRseq Mature CNMF subtypes'

Cluster Labels	1	2	3	4	5	6	7
Number of samples	44	54	71	18	53	82	24

'MIRseq Mature CNMF subtypes' versus 'Time to Death'

P value = 0.184 (logrank test), Q value = 0.39

Table S106. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	340	120	0.0 - 120.8 (19.8)
subtype1	43	19	0.4 - 108.8 (18.5)
subtype2	54	25	0.2 - 120.8 (20.7)
subtype3	68	27	0.3 - 114.3 (18.2)
subtype4	18	6	1.0 - 83.2 (25.5)
subtype5	52	14	0.0 - 107.1 (21.9)
subtype6	82	19	0.3 - 90.5 (19.4)
subtype7	23	10	1.5 - 76.4 (21.8)

Figure S97. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq Mature CNMF subtypes' versus 'YEARS_TO_BIRTH'

P value = 3.89e-06 (Kruskal-Wallis (anova)), Q value = 0.00016

Table S107. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	342	59.2 (13.1)
subtype1	43	63.1 (9.3)
subtype2	53	62.0 (11.3)
subtype3	71	52.0 (15.2)
subtype4	17	63.5 (13.5)
subtype5	53	57.8 (12.9)
subtype6	81	62.8 (10.3)
subtype7	24	56.1 (15.9)

Figure S98. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'MIRseq Mature CNMF subtypes' versus 'PATHOLOGIC_STAGE'

P value = 0.0598 (Fisher's exact test), Q value = 0.18

Table S108. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV	STAGE IVA	STAGE IVB
ALL	157	83	3	59	7	8	2	1	2
subtype1	15	13	0	9	3	2	0	0	1
subtype2	22	10	1	12	3	1	0	0	1
subtype3	36	12	1	13	1	3	1	0	0
subtype4	9	0	0	2	0	1	0	0	0
subtype5	22	17	0	6	0	1	1	1	0
subtype6	43	24	0	11	0	0	0	0	0
subtype7	10	7	1	6	0	0	0	0	0

Figure S99. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'MIRseq Mature CNMF subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.0258 (Fisher's exact test), Q value = 0.088

Table S109. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	167	91	73	12
subtype1	16	14	10	4
subtype2	22	13	16	3
subtype3	38	14	17	2
subtype4	11	0	4	1
subtype5	25	18	8	2
subtype6	45	25	11	0
subtype7	10	7	7	0

Figure S100. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'MIRseq Mature CNMF subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.509 (Fisher's exact test), Q value = 0.67

Table S110. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	238	4
subtype1	27	1
subtype2	37	0
subtype3	54	2
subtype4	8	0
subtype5	30	1
subtype6	60	0
subtype7	22	0

Figure S101. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'MIRseq Mature CNMF subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.631 (Fisher's exact test), Q value = 0.76

Table S111. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	255	4
subtype1	36	1
subtype2	34	1
subtype3	56	1
subtype4	10	0
subtype5	32	1
subtype6	66	0
subtype7	21	0

Figure S102. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'MIRseq Mature CNMF subtypes' versus 'GENDER'

P value = 0.0295 (Fisher's exact test), Q value = 0.098

Table S112. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	115	231
subtype1	13	31
subtype2	20	34
subtype3	31	40
subtype4	9	9
subtype5	17	36
subtype6	16	66
subtype7	9	15

Figure S103. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #7: 'GENDER'

'MIRseq Mature CNMF subtypes' versus 'RADIATION_THERAPY'

P value = 0.72 (Fisher's exact test), Q value = 0.8

Table S113. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	315	9
subtype1	41	1
subtype2	52	1
subtype3	64	1
subtype4	17	0
subtype5	47	1
subtype6	72	5
subtype7	22	0

Figure S104. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

'MIRseq Mature CNMF subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 0.00539 (Fisher's exact test), Q value = 0.027

Table S114. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

nPatients	FIBROLAMELLAR CARCINOMA	HEPATOCELLULAR CARCINOMA	HEPATOCHOLANGIOCARCINOMA (MIXED)
ALL	2	337	7
subtype1	0	42	2
subtype2	0	54	0
subtype3	0	71	0
subtype4	0	18	0
subtype5	2	47	4
subtype6	0	82	0
subtype7	0	23	1

Figure S105. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

'MIRseq Mature CNMF subtypes' versus 'RESIDUAL_TUMOR'

P value = 0.395 (Fisher's exact test), Q value = 0.57

Table S115. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	303	17	1	19
subtype1	40	1	0	1
subtype2	45	2	0	5
subtype3	63	3	0	5
subtype4	18	0	0	0
subtype5	40	7	1	4
subtype6	75	3	0	3
subtype7	22	1	0	1

Figure S106. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

'MIRseq Mature CNMF subtypes' versus 'RACE'

P value = 0.00046 (Fisher's exact test), Q value = 0.005

Table S116. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #11: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	2	155	16	163
subtype1	0	14	1	27
subtype2	0	17	7	30
subtype3	1	44	0	25
subtype4	0	3	2	12
subtype5	0	21	2	28
subtype6	1	42	3	32
subtype7	0	14	1	9

Figure S107. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #11: 'RACE'

'MIRseq Mature CNMF subtypes' versus 'ETHNICITY'

P value = 0.973 (Fisher's exact test), Q value = 0.97

Table S117. Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	16	312
subtype1	2	38
subtype2	4	46
subtype3	3	65
subtype4	0	17
subtype5	2	48
subtype6	4	75
subtype7	1	23

Figure S108. Get High-res Image Clustering Approach #9: 'MIRseq Mature CNMF subtypes' versus Clinical Feature #12: 'ETHNICITY'

Clustering Approach #10: 'MIRseq Mature cHierClus subtypes'

Table S118. Description of clustering approach #10: 'MIRseq Mature cHierClus subtypes'

Cluster Labels	1	2	3	4	5	6
Number of samples	132	50	73	40	47	4

'MIRseq Mature cHierClus subtypes' versus 'Time to Death'

P value = 0.09 (logrank test), Q value = 0.25

Table S119. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

	nPatients	nDeath	Duration Range (Median), Month
ALL	340	120	0.0 - 120.8 (19.8)
subtype1	128	55	0.3 - 114.3 (20.0)
subtype2	50	18	0.3 - 102.7 (21.9)
subtype3	72	20	0.2 - 90.3 (18.9)
subtype4	39	17	0.0 - 120.8 (15.5)
subtype5	47	9	0.3 - 108.8 (19.2)
subtype6	4	1	10.3 - 75.7 (23.2)

Figure S109. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #1: 'Time to Death'

'MIRseq Mature cHierClus subtypes' versus 'YEARS_TO_BIRTH'

P value = 5.99e-05 (Kruskal-Wallis (anova)), Q value = 9e-04

Table S120. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

	nPatients	Mean (Std.Dev)
ALL	342	59.2 (13.1)
subtype1	130	58.1 (13.2)
subtype2	49	64.2 (10.3)
subtype3	72	53.5 (15.2)
subtype4	40	62.4 (10.5)
subtype5	47	62.9 (10.9)
subtype6	4	65.2 (9.0)

Figure S110. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #2: 'YEARS_TO_BIRTH'

'MIRseq Mature cHierClus subtypes' versus 'PATHOLOGIC_STAGE'

P value = 0.113 (Fisher's exact test), Q value = 0.27

Table S121. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

nPatients	STAGE I	STAGE II	STAGE III	STAGE IIIA	STAGE IIIB	STAGE IIIC	STAGE IV	STAGE IVA	STAGE IVB
ALL	157	83	3	59	7	8	2	1	2
subtype1	54	31	0	28	3	1	1	0	1
subtype2	26	12	1	3	2	1	0	0	0
subtype3	31	20	2	10	0	5	1	1	0
subtype4	20	5	0	10	2	1	0	0	1
subtype5	23	15	0	7	0	0	0	0	0
subtype6	3	0	0	1	0	0	0	0	0

Figure S111. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #3: 'PATHOLOGIC_STAGE'

'MIRseq Mature cHierClus subtypes' versus 'PATHOLOGY_T_STAGE'

P value = 0.366 (Fisher's exact test), Q value = 0.57

Table S122. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

nPatients	T1	T2	T3	T4
ALL	167	91	73	12
subtype1	59	34	34	4
subtype2	28	13	6	3
subtype3	34	23	13	3
subtype4	20	6	12	2
subtype5	24	15	7	0
subtype6	2	0	1	0

Figure S112. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #4: 'PATHOLOGY_T_STAGE'

'MIRseq Mature cHierClus subtypes' versus 'PATHOLOGY_N_STAGE'

P value = 0.384 (Fisher's exact test), Q value = 0.57

Table S123. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

nPatients	0	1
ALL	238	4
subtype1	83	1
subtype2	33	0
subtype3	55	3
subtype4	29	0
subtype5	35	0
subtype6	3	0

Figure S113. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #5: 'PATHOLOGY_N_STAGE'

'MIRseq Mature cHierClus subtypes' versus 'PATHOLOGY_M_STAGE'

P value = 0.751 (Fisher's exact test), Q value = 0.82

Table S124. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

nPatients	0	1
ALL	255	4
subtype1	96	2
subtype2	34	0
subtype3	59	1
subtype4	25	1
subtype5	39	0
subtype6	2	0

Figure S114. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #6: 'PATHOLOGY_M_STAGE'

'MIRseq Mature cHierClus subtypes' versus 'GENDER'

P value = 0.0388 (Fisher's exact test), Q value = 0.13

Table S125. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

nPatients	FEMALE	MALE
ALL	115	231
subtype1	40	92
subtype2	16	34
subtype3	34	39
subtype4	14	26
subtype5	9	38
subtype6	2	2

Figure S115. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #7: 'GENDER'

'MIRseq Mature cHierClus subtypes' versus 'RADIATION_THERAPY'

P value = 0.15 (Fisher's exact test), Q value = 0.33

Table S126. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

nPatients	NO	YES
ALL	315	9
subtype1	117	4
subtype2	48	0
subtype3	68	1
subtype4	38	0
subtype5	40	4
subtype6	4	0

Figure S116. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #8: 'RADIATION_THERAPY'

'MIRseq Mature cHierClus subtypes' versus 'HISTOLOGICAL_TYPE'

P value = 0.206 (Fisher's exact test), Q value = 0.41

Table S127. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

nPatients	FIBROLAMELLAR CARCINOMA	HEPATOCELLULAR CARCINOMA	HEPATOCHOLANGIOCARCINOMA (MIXED)
ALL	2	337	7
subtype1	0	129	3
subtype2	0	50	0
subtype3	2	67	4
subtype4	0	40	0
subtype5	0	47	0
subtype6	0	4	0

Figure S117. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #9: 'HISTOLOGICAL_TYPE'

'MIRseq Mature cHierClus subtypes' versus 'RESIDUAL_TUMOR'

P value = 0.588 (Fisher's exact test), Q value = 0.72

Table S128. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

nPatients	R0	R1	R2	RX
ALL	303	17	1	19
subtype1	115	8	1	6
subtype2	45	1	0	4
subtype3	62	4	0	7
subtype4	32	3	0	2
subtype5	45	1	0	0
subtype6	4	0	0	0

Figure S118. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #10: 'RESIDUAL_TUMOR'

'MIRseq Mature cHierClus subtypes' versus 'RACE'

P value = 0.387 (Fisher's exact test), Q value = 0.57

Table S129. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #11: 'RACE'

nPatients	AMERICAN INDIAN OR ALASKA NATIVE	ASIAN	BLACK OR AFRICAN AMERICAN	WHITE
ALL	2	155	16	163
subtype1	1	61	3	63
subtype2	0	18	4	23
subtype3	0	38	3	32
subtype4	0	15	5	20
subtype5	1	22	1	22
subtype6	0	1	0	3

Figure S119. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #11: 'RACE'

'MIRseq Mature cHierClus subtypes' versus 'ETHNICITY'

P value = 0.919 (Fisher's exact test), Q value = 0.93

Table S130. Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #12: 'ETHNICITY'

nPatients	HISPANIC OR LATINO	NOT HISPANIC OR LATINO
ALL	16	312
subtype1	8	117
subtype2	2	43
subtype3	2	68
subtype4	2	37
subtype5	2	43
subtype6	0	4

Figure S120. Get High-res Image Clustering Approach #10: 'MIRseq Mature cHierClus subtypes' versus Clinical Feature #12: 'ETHNICITY'

Methods & Data

Input

Cluster data file = /xchip/cga/gdac-prod/tcga-gdac/jobResults/GDAC_mergedClustering/LIHC/26874120/LIHC.mergedcluster.txt
Clinical data file = /xchip/cga/gdac-prod/tcga-gdac/jobResults/Append_Data/LIHC/26850638/LIHC.merged_data.txt
Number of patients = 377
Number of clustering approaches = 10
Number of selected clinical features = 12
Exclude small clusters that include fewer than K patients, K = 3

Clustering approaches

CNMF clustering

consensus non-negative matrix factorization clustering approach (Brunet et al. 2004)

Consensus hierarchical clustering

Resampling-based clustering method (Monti et al. 2003)

Survival analysis

For survival clinical features, the Kaplan-Meier survival curves of tumors with and without gene mutations were plotted and the statistical significance P values were estimated by logrank test (Bland and Altman 2004) using the 'survdiff' function in R

Fisher's exact test

For binary clinical features, two-tailed Fisher's exact tests (Fisher 1922) were used to estimate the P values using the 'fisher.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] Brunet et al., Metagenes and molecular pattern discovery using matrix factorization, PNAS 101(12):4164-9 (2004)

[2] Monti et al., Consensus Clustering: A Resampling-Based Method for Class Discovery and Visualization of Gene Expression Microarray Data, Machine Learning 52(1):91-118 (2003)

[3] Bland and Altman, Statistics notes: The logrank test, BMJ 328(7447):1073 (2004)

[4] Fisher, R.A., On the interpretation of chi-square from contingency tables, and the calculation of P, Journal of the Royal Statistical Society 85(1):87-94 (1922)

[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)

Made with Nozzle