disclaimer - seoul national...

74
저작자표시-비영리-변경금지 2.0 대한민국 이용자는 아래의 조건을 따르는 경우에 한하여 자유롭게 l 이 저작물을 복제, 배포, 전송, 전시, 공연 및 방송할 수 있습니다. 다음과 같은 조건을 따라야 합니다: l 귀하는, 이 저작물의 재이용이나 배포의 경우, 이 저작물에 적용된 이용허락조건 을 명확하게 나타내어야 합니다. l 저작권자로부터 별도의 허가를 받으면 이러한 조건들은 적용되지 않습니다. 저작권법에 따른 이용자의 권리는 위의 내용에 의하여 영향을 받지 않습니다. 이것은 이용허락규약 ( Legal Code) 을 이해하기 쉽게 요약한 것입니다. Disclaimer 저작자표시. 귀하는 원저작자를 표시하여야 합니다. 비영리. 귀하는 이 저작물을 영리 목적으로 이용할 수 없습니다. 변경금지. 귀하는 이 저작물을 개작, 변형 또는 가공할 수 없습니다.

Upload: others

Post on 14-Jul-2020

0 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

저 시-비 리- 경 지 2.0 한민

는 아래 조건 르는 경 에 한하여 게

l 저 물 복제, 포, 전송, 전시, 공연 송할 수 습니다.

다 과 같 조건 라야 합니다:

l 하는, 저 물 나 포 경 , 저 물에 적 된 허락조건 명확하게 나타내어야 합니다.

l 저 터 허가를 면 러한 조건들 적 되지 않습니다.

저 에 른 리는 내 에 하여 향 지 않습니다.

것 허락규약(Legal Code) 해하 쉽게 약한 것 니다.

Disclaimer

저 시. 하는 원저 를 시하여야 합니다.

비 리. 하는 저 물 리 목적 할 수 없습니다.

경 지. 하는 저 물 개 , 형 또는 가공할 수 없습니다.

Page 2: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

학 사 학 논

Gene Expression Profiling of Breast

Cancer according to

Mammographic Microcalcifications

미 회 에 른

암 자 프 일링

2016 08 월

울 학 학원

임상 과학과

신 승

Page 3: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

i

Abstract

Introduction: Mammographic microcalcifications are important in early

detection and diagnosis of breast cancer. However, the underlying biological

mechanisms of microcalcifications and the association between

microcalcifications and prognosis of breast cancer are largely unknown. This

study was to investigate association between mammographic

microcalcifications and gene expression profiles of breast cancer using

microarray analysis.

Methods: Gene expression analysis was performed using Affymetrix

GeneChip® Human Gene 2.0 ST arrays in 168 breast cancer patients (median

age, 50.0 years; range, 21-79 years). Mammographic microcalcifications of

these patients were reviewed by three radiologists and were grouped into no

microcalcification (n=99), low-to-intermediate (n=37) and highly (n=32)

suspicious microcalcifications group. To identify differentially expressed

genes (DEGs) between three groups, the one-way analysis of variance was

carried out with post hoc comparisons made with Tukey's honest significant

difference test and P value < 0.05 was applied. To explore biological meaning

behind DEGs, we used DAVID for gene ontology analysis and BioLattice for

clustering analysis.

Results: Total 2551 genes showed different expressions in comparison of

three groups; 1838 DEGs (955 up-regulated and 883 down-regulated) in

highly suspicious microcalcifications/no microcalcification comparison and

484 DEGs (342 up-regulated and 142 down-regulated) in highly

suspicious/low-to-intermediate microcalcifications comparison, 457 DEGs

Page 4: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

ii

(126 up-regulated and 331 down-regulated) in low-to-intermediate/no

microcalcification comparison were discovered. GO analysis revealed that

immune, defense and inflammatory response were decreased in highly

suspicious microcalcifications group when it compared to no

microcalcification group. Clustering analysis using BioLattice also discovered

that immune system is associated with mammographic microcalcifications in

total population and human epidermal growth factor receptor 2-negative

subgroup.

Conclusions: Gene expression patterns are different according to the status of

mammographic microcalcifications in breast cancer. Breast cancers with

mammographic microcalcifications are associated with decreased immune

system activity.

Keywords: Breast Neoplasms, Microcalcification, Microarray analysis,

Human epidermal growth factor receptor 2, Immune system

Student number: 2014-30911

Page 5: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

iii

Contents

Abstract ----------------------------------------------------------- i

Contents ---------------------------------------------------------- iii

Lists for Tables and Figures ------------------------------------ iv

List of Abbreviations -------------------------------------------- v

Introduction ------------------------------------------------------- 1

Materials and Methods ------------------------------------------ 3

Results ------------------------------------------------------------- 9

Discussion -------------------------------------------------------- 27

References -------------------------------------------------------- 32

Appendix 1 ------------------------------------------------------- 37

Appendix 2 ------------------------------------------------------- 44

Appendix 3 ------------------------------------------------------- 47

Appendix 4 ------------------------------------------------------- 58

Appendix 5 ------------------------------------------------------- 63

초록 (국문) ------------------------------------------------------ 66

Page 6: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

iv

Lists for Tables and Figures

Table 1 ------------------------------------------------------------ 10

Table 2 ------------------------------------------------------------ 13

Figure 1 ----------------------------------------------------------- 3

Figure 2 ----------------------------------------------------------- 4

Figure 3 ----------------------------------------------------------- 16

Figure 4 ----------------------------------------------------------- 17

Figure 5 ----------------------------------------------------------- 20

Figure 6 ----------------------------------------------------------- 22

Figure 7 ----------------------------------------------------------- 24

Figure 8 ----------------------------------------------------------- 26

Page 7: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

v

List of Abbreviations

ANOVA = analysis of variance

DAVID = Database for Annotation, Visualization, and Integrated Discovery

DCIS = ductal carcinoma in situ

DEG = differentially expressed gene

ER = estrogen receptor

GO = gene ontology

HER2 = human epidermal growth factor receptor 2

HR = hormone receptor

PR = progesterone receptor

RNA = ribonucleic acid

Page 8: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

1

Introduction

Mammography is established screening tool for breast cancer (1) and

microcalcifications are one of the most important findings for detection of

breast cancer. Tabar et al. and others reported that cancers with

microcalcifications of different morphology had different outcome (2-5).

Evidences suggest that microcalcifications affect the prognosis of breast

cancer, in other words, breast cancers with microcalcifications are thought to

be more aggressive and have poorer prognosis than those without

microcalcifications (2, 6-11).

Previous studies have attempted to evaluate and interpret the clinical

perspective of the relationship between mammographic microcalcifications

and the expression of selected biological markers, such as estrogen receptor

(ER), progesterone receptor (PR) or human epidermal growth factor receptor

2 (HER2) using immunohistochemistry. Although there are several reports

that breast cancers with mammographic microcalcifications are more

frequently associated with HER2 overexpression (12-15), those still remain

unclear that what the underlying molecular mechanism regarding formation of

microcalcifications is, how HER2 overexpression affect morphology of

microcalcifications, and how microcalcifications are associated with

prognosis.

Owing to recent advances in computer science, genomics and

imaging technique, there has been growing interests in exploring multiscale

relationships of human breast cancers (16-19). Linking the multiscale

Page 9: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

2

relationships between clinical, imaging, pathologic and genomic data are so

called radiogenomics. Identifying biological background associated with

progression and tumor aggressiveness in conjunction with imaging features

such as microcalcifications would help to make treatment decisions (20).

However, in breast cancer, mainstay of radiogenomic study often uses MR

imaging. Thus the relationships between mammographic imaging findings

and global transcriptomic profiles are overlooked and remain to be examined.

Accordingly, we hypothesized that gene expression profiles of breast cancer

would be different according to microcalcifications status. The purpose of this

study was to investigate association between mammographic

microcalcifications and gene expression profiles of breast cancer using

microarray analysis.

Page 10: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

3

Materials and Methods

Study population

This retrospective study was approved by the institutional review

board of Seoul National University Hospital (IRB No. 1409-128-612) and all

patients provided written informed consent for their breast cancer tissue to be

used for genome sequencing (IRB No. 1405-088-580) before operation. Of

the 182 patients enrolled, 14 patients were excluded as non-invasive cancer (n

= 4), or no initial digital mammography (n = 5), or excisional biopsy (n = 2),

or neoadjuvant chemotherapy (n = 1), or metastasis on initial presentation (n =

2). Thus, 168 breast cancer women (mean age, 50.7 years; range, 21-79 years)

were finally included for analysis.

Figure 1 Schematic of inclusion and exclusion criteria and selection process

of this study cohort

Page 11: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

4

Radiologic review

By using the microcalcifications at mammography as a criterion for

grouping the patients, three radiologists with different degrees of experience

in interpreting mammography independently analyzed microcalcifications

without access to genomic data. After each radiologist had finished the

analysis, a consensus was established. Finally, the population was stratified on

the basis of malignancy probability into no microcalcification (n = 99), low-

to-intermediate (n = 37) and highly (n = 32) suspicious groups.

Figure 2. Examples of microcalcifications. A, Highly suspicious

microcalcifications. B, Low-to-intermediate microcalcifications

Pathologic review

Hematoxylin and eosin–stained slides of frozen human tumor tissue

were examined per standard protocol for pathologic diagnosis.

Immunohistochemical analysis was performed on formalin-fixed, paraffin-

Page 12: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

5

embedded 4-mm tissue sections using the primary mouse monoclonal

antibodies for ER, PR, HER2, respectively. For equivocal HER2 results (2+),

the status was determined by means of fluorescence in-situ hybridization (21).

Tissue samples and microarray analysis

Tissue samples were dissected through the centers of carcinomatous

region during surgery at our hospital between 2003-2012. Those were frozen

in liquid nitrogen within 20 min following surgical devascularization and

stored at -80˚C. Total RNA from each sample was extracted using TRIZOL

reagent. RNA quality was assessed by Agilent 2100 bioanalyser using RNA

6000 Nano Chip (Agilent Technologies, Amstelveen, The Netherlands), and

quantity was determined by ND-2000 Spectrophotometer (Thermo Inc., DE,

USA). The median RNA extracted was 1.202 g/L (range 0.143-2.986 g/L).

Total RNA was measured by determining UV absorbance at 260nm. Purity of

sample was assessed by measuring OD 260:280nm and OD 260:230nm. The

integrity of RNA samples was confirmed by the appearance of distinct 28S

and18S bands of ribosomal RNA. RNA integrity number (RIN) was

determined using the RIN algorithm of the Agilient 2100 expert software (22).

The quality of the RNA was good with the standard 260/280 ratio and

260/230 ratio of absorbance greater than 1.7 and 1.3 per sample, respectively.

Mean 28S/18S ratio was 1.0 (range 0.3-1.9) and RIN was greater than or

equal to 5.0. In this study, we performed global gene expression analysis

using Affymetrix GeneChip® Human Gene 2.0 ST oligonucleotide arrays

Page 13: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

6

(53,617 probes). Affymetrix Model 3000 G7 scanner and Affymetrix

Command Console software 1.1 were used for scanning and data extraction.

The raw excel file containing intensity data was used for further analysis. For

normalization, Robust Multiarray-Average algorithm was used (23) which

was developed in the Speed Lab at UC Berkeley.

Statistical analysis and bioinformatics analysis

Demographic characteristics, clinical, pathologic, and

mammographic findings were compared between groups using chi-square test

and Fisher’s exact test for categorical variables. One-way analysis of variance

(ANOVA) was performed for numerical variables.

Statistical analyses of microarray data were performed using R

software, version 3.2.4 (http://www.r‐project.org/). R oligo package was used

for processing microarray data, which is freely available on the World Wide

Web (http://www. bioconductor.org) (24, 25). To identify differentially

expressed genes between three groups the One-way ANOVA was carried out

with post hoc comparisons made with Tukey's honest significant difference

test (26) and P value < 0.05 were applied for subsequent data analysis.

To gain insight into the underlying biology of DEGs related to

mammographic microcalcifications, functional categories enriched in the

differentially expressed genes were identified using the functional annotation

and clustering tool of the Database for Annotation, Visualization, and

Integrated Discovery (DAVID) v6.7 (https://david.ncifcrf.gov/) (27, 28). The

Page 14: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

7

probability that a GO biological process term is overrepresented was

determined by a modified Fisher’s exact test, comparing the proportion of

genes in the whole genome which are part of that GO term, to the proportion

of the differentially expressed genes which are part of the same GO term, and

was expressed as an Expression Analysis Systematic Explorer (EASE) score

(29). In addition, to interpret and organize observed biological changes, we

used BioLattice (http://www.snubi.org/software/biolattice/), a mathematical

framework based on concept lattice analysis to make association of gene

expression clusters with biological ontologies or biological pathways after k-

means clustering method with respect to the Pearson correlation similarity

measure. We set arbitrary k to a cluster contain genes from 15 to 30. We

chose threshold of P < 0.001 (30). BioLattice considers gene expression

clusters as objects and annotations as attributes and provide a graphical

summary of the order relations by arranging them on a concept lattice in an

order based on set inclusion relation. Rather than interpreting one cluster at a

time, BioLattice integrates all gene expression clusters and annotations into a

unified framework: a lattice of concepts. The top element of a lattice is a unit

concept, representing a concept that contains all objects. The bottom element

is a zero concept having no object. Core–periphery analysis decomposes

BioLattice into four disjoint core–periphery substructures: ‘core’,

‘communicating’, ‘peripheral’ and ‘independent’. The ‘core’ sub-lattice is

defined as the maximal atom sub-lattice in terms of size (i.e., the number of

the red colored concepts). The set of all lower bounds to the nodes included in

the ‘core’ sub-lattice (excluding those included in the ‘core’) is defined as the

Page 15: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

8

‘communicating’ substructure (in green). When an atom equals a coatom, we

call it ‘independent’ (in yellow). All the rest are defined as ‘peripheral’ (in

gray).

To identify the possible influence of HER2 positivity on

mammographic microcalcifications, correlation analysis between HER2 score

(0, 1+, 2+, and 3+) and microcalcifications (no, low-to-intermediate, and

highly suspicious) and between HER2 score (0, 1+, 2+, and 3+) and gene

expressions were carried out. Correlation analysis between microcalcifications

(no, low-to-intermediate, and highly suspicious) and gene expressions were

performed as well. In addition, we performed subgroup analysis according to

the status of HER2 (positive or negative).

Page 16: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

9

Results

Clinicopathological data

Median patient age was 50.0 years. Mean tumor size for the study

group was 3.1cm. The patients had invasive ductal carcinomas (89.9%) or

invasive lobular carcinoma (1.8%) or others with a clinical stage of I (14.3%),

II (63.1%), or III (22.6%). Ninety two patients were hormone receptor (HR)

positive and the other 76 were HR negative. As for HER2 receptor status, 38

patients were positive and 130 were negative. Detailed demographic

characteristics and clinicopathologic findings were summarized in Table 1.

HER2 positivity, ductal carcinoma in situ (DCIS) and comedo necrosis (all P

< 0.001) were more frequently seen in the highly suspicious

microcalcifications group. The other demographic, clinical and pathologic

findings are not significantly different between three groups.

Page 17: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

10

Table 1. Characteristics of patients

No

microcalcifications (n = 99)

Low-to-intermediate

microcalcifications (n = 37)

Highly suspicious microcalcifications

(n = 32) P value

Age, years 0.383 ≤50 53 (53.5) 16 (43.2) 19 (59.4) >50 46 (46.5) 21 (56.8) 13 (40.6)

Menopausal status 0.191 Premenopausal 51 (51.5) 13 (35.1) 18 (56.3) Postmenopausal 48 (48.5) 23 (62.2) 14 (43.8)

Clinical symptom * 0.151 Yes 91 (91.9) 30 (81.1) 27 (84.4) No 8 (8.1) 7 (18.9) 5 (15.6)

Mean tumor size ± SD (cm)† 3.2 ± 1.2 3.1 ± 1.3 2.7 ± 2.1 0.179 No. of nodal metastasis

0 56 (56.6) 18 (48.7) 22 (68.8) 1 or more 43 (43.4) 19 (51.4) 10 (31.3) 0.239 ≥ 4 or more 19 (19.2) 10 (27.0) 6 (18.8) 0.575

Histologic grade 0.799 I-II 31 (31.3) 10 (27.0) 11 (34.4) III 68 (68.7) 27 (73.0) 21 (65.6)

HR positivity 55 (55.6) 22 (59.5) 15 (46.9) 0.560 HER2 positivity 12 (12.1) 10 (27.0) 16 (50.0) <0.001 Triple-negative 33 (33.3) 11 (29.8) 5 (15.6) 0.159 DCIS 58 (58.6) 28 (75.7) 31 (96.9) <0.001 Comedo necrosis 23 (23.2) 21 (56.8) 29 (90.6) <0.001

Page 18: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

11

HR, hormone receptor. HER2, human epidermal growth factor receptor 2. DCIS, ductal carcinoma in situ. Unless otherwise indicated, data are number of patients and data in parentheses are percentages. * P value was calculated by Fisher’s exact. Others were calculated by chi-square test.

†Numbers represent the mean values ± standard deviations. P value was calculated by ANOVA.

Page 19: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

12

Imaging analysis

Mammographic features of the patients are shown in Table 2. In

highly suspicious microcalcifications group, majority of patients presented as

irregular mass with microcalcifications (P < 0.001). There was no significant

difference between three groups regarding breast composition or margin and

density of mass. All highly suspicious microcalcifications were classified

according Breast Imaging-Reporting and Data System (BI-RADS) guidelines;

morphology of microcalcifications as coarse heterogeneous (n = 1), fine

pleomorphic (n = 21), fine linear and linear branching (n = 10); distribution as

regional (n = 1), grouped (n = 12), and segmental (n = 19).

Page 20: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

13

Table 2. Mammographic findings of patients

No

microcalcifications (n = 99)

Low-to-intermediate

microcalcifications (n = 37)

Highly suspicious microcalcifications

(n = 32) P value

Breast composition 0.122 Fatty 34 (34.3) 10 (27.0) 5 (15.6) Dense 65 (65.7) 27 (73.0) 27 (84.4)

Mammographic finding <0.001 Mass with calcification 0 (0) 36 (97.3) 28 (87.5) Mass without calcification 88 (88.9) 0 (0) 0 (0) Asymmetry or architectural

distortion 11 (11.1) 1 (2.7) 4 (12.5)

Mass shape* 0.003 Oval or round 26 (29.5) 12 (33.3) 0 (0) Irregular 62 (70.5) 24 (66.7) 28 (100.0)

Mass margin* † 0.141 Circumscribed 11 (12.5) 4 (11.1) 0 (0) Not circumscribed 77 (87.5) 32 (88.9) 28 (100.0)

Mass density* † 0.132 High 77 (87.5) 30 (83.3) 24 (85.7) Equal 11 (12.5) 6 (16.7) 4 (14.3)

Data are number of patients and data in parentheses are percentages. * Shape, margin and density of mass was evaluated only in mass cases.

† P value was calculated by Fisher’s exact. Others were calculated by chi-square test.

Page 21: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

14

Imaging-genomic correlation

When we compared the genomic composition of three groups, 2551

genes were differentially expressed at level of P < 0.05; 1838 DEGs (955 up

and 883 down) in highly suspicious/no microcalcifications group comparison

set, 484 DEGs (342 up and 142 down) in highly suspicious/low-to-

intermediate group comparison set, and 457 DEGs (126 up and 331 down) in

low-to-intermediate group/no microcalcifications group comparison set were

extracted. List of DEGs (top 200 genes are shown, ordered by P value) in

highly suspicious microcalcifications/no microcalcification comparison set are

shown in Appendix 1. With these DEG sets, gene ontology analyses and

functional annotation clustering were performed with DAVID. Focusing of

DEGs obtained from highly suspicious/no microcalcifications comparison set,

we found that the 10 top-ranked biological functions (P value from 10-23 to 10-

8, Figure 3) included immune response, antigen processing and presentation,

defense response, regulation of cytokine production, positive regulation of

immune system and response to wounding. Results of functional annotation

clustering are summarized in Appendix 2. To find out where these GO terms

in biological process category came from, up-regulated genes and down-

regulated genes are separately analyzed. We discovered that these immune-

related GO terms are mostly from down-regulated DEGs.

In addition, we used Pearson correlation as similarity measure and

grouped genes into 60, 17, and 16 clusters using k-means clustering (31) for

BioLattice in comparison of highly suspicious/no microcalcification, highly

suspicious/low-to-intermediate, and low-to-intermediate group/no

Page 22: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

15

microcalcification, respectively. We set arbitrary k as 60, 17 and 16 to each

cluster contain 15-30 genes. We could discover several annotation clusters

associated with immune system in highly suspicious/no microcalcification

comparison set by clustering analysis using BioLattice. Figure 4 shows lattice

of concepts, constructed from the highly suspicious/no microcalcification

comparison set (n = 1838) in total population having 60 clusters annotated by

GO terms in the biological process category. Only 24 among 60 clusters

demonstrate at least one significant GO term(s) (P < 0.001). Overall, the

dataset shows 125 significant annotations with 106 unique GO terms. Four

core concepts (red color) are associated with immune system including

defense response, immune response, and inflammatory response. The upper

semicircle of a concept node contains the concept ID and the lower semicircle

the list of the cluster IDs belonging to the node.

Page 23: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

16

Figure 3. Enriched gene ontology biologic process terms using DEGs (n = 1838) from highly suspicious microcalcifications/no

microcalcification comparison set. Immune system-related GO terms in biological process category are enriched in no microcalcification

group.

Page 24: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

17

Figure 4. Concept lattice constructed from the highly suspicious/no microcalcification comparison set (n = 1838) in total population having

60 clusters annotated by GO terms in the biological process category. Only 24 among 60 clusters demonstrate at least one significant GO

term(s) (P < 0.001). Overall, the dataset shows 125 significant annotations with 106 unique GO terms. The core–periphery substructures

marked with colors (i.e., core in red, communicating in green, independent in yellow and peripheral in gray).

Page 25: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

18

Before interpreting the results, because we know that HER2

overexpression is correlated with mammographic microcalcifications; we had

concern that the composition of HER2 overexpression between groups is

significantly different (P < 0.001). Thus, we decided to analyze its correlation

with genomic composition and microcalcifications score and to conduct

further subgroup analysis according to the status of HER2 positivity.

In analysis of correlation between HER2 score (0, 1+, 2+, 3+) and

microcalcifications (no, low-to-intermediate, highly suspicious), there was

weak correlation with correlation coefficient (r) 0.264 and P value < 0.001. In

analysis of correlation between HER2 score (0, 1+, 2+, 3+) and genomic

composition, 297 genes are significantly correlated with HER2 score with

Bonferroni-adjusted P value < 0.05 (Appendix 3). As expected, ERBB2 (r =

0.716) and GRB7 (r = 0.652) are ranked first and second in order of r. When

those 297 genes applied to GO database, there was no significant GO term

with P value < 0.05. Between these 297 genes and gene set initially obtained

from highly suspicious/no microcalcifications comparison set, only 62 genes

(20.9%) were overlapped. In addition, we carried out correlation analysis

between microcalcifications (no, low-to-intermediate, highly suspicious) and

genomic composition and obtained 116 genes with r > 0.3 and false discovery

rate < 0.05 (Appendix 4). Interestingly, there are several other genes with

larger correlation coefficient than ERBB2. It suggests that there may be

pathways other than HER2 signaling pathway of microcalcifications

formation in the breast.

In HER2 positive subgroup (n = 38) with same methodology above,

Page 26: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

19

1044 DEGs were extracted at P value < 0.05; 187 DEGs (91 up and 96 down)

in highly suspicious/no microcalcification comparison set, 474 DEGs (188 up

and 286 down) in highly suspicious/low-to-intermediate comparison set, and

498 DEGs (374 up and 124 down) in low-to-intermediate/no

microcalcification comparison set. There was no significant GO term in

biological process category or cluster performed using DEGs in all

comparison set.

In HER2 negative subgroup (n = 130), total 1679 DEGs were

extracted at P value < 0.05; 1047 DEGs (465 up and 582 down) in highly

suspicious/no microcalcification comparison set, 199 DEGs (123 up and 76

down) in highly suspicious/low-to-intermediate comparison set, and 370

DEGs (130 up and 240 down) in low-to-intermediate/no microcalcification

comparison set. Focusing on DEGs obtained from highly suspicious/no

microcalcification comparison set, we found that the 10 top-ranked biological

functions (P value from 10-39 to 10-9, Figure 5) included immune response,

defense response, antigen processing and presentation, inflammatory response,

response to wounding, and regulation of cytokine production. Results of

functional annotation clustering are summarized in Appendix 5.

Page 27: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

20

Figure 5. Enriched gene ontology biologic process terms using DEGs (n = 1047) from highly suspicious/no microcalcification comparison set

in HER2-negative subgroup.

Page 28: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

21

In addition, we applied k-means clustering (31) to construct 20 and

40 groups of coexpressed genes for BioLattice for HER2-positive and HER2-

negative group, respectively. In HER2-positive subgroup, no statistically

significant cluster was discovered. In concept lattice constructed from the

highly suspicious/no microcalcification comparison set (n = 1047) in HER2-

negative subgroup having 40 clusters annotated by GO terms in the biological

process category. Only 16 among 40 clusters demonstrate at least one

significant GO term(s) (P < 0.001). Overall, the dataset shows 54 significant

annotations with 43 unique GO terms. Four core concepts (red color) and

three of them are associated with immune system including defense response,

immune response, and inflammatory response (Figure 6).

Page 29: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

22

Figure 6. Concept lattice constructed from the highly suspicious/no microcalcification comparison set (n = 1047) in HER2-negative subgroup

having 40 clusters annotated by GO terms in the biological process category. Only 16 among 40 clusters demonstrate at least one significant

GO term(s) (P < 0.001). Overall, the dataset shows 54 significant annotations with 43 unique GO terms. The core–periphery substructures

marked with colors (i.e., core in red, communicating in green, independent in yellow and peripheral in gray).

Page 30: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

23

In HER2-negative subgroup, further subgroup analysis was done

according to HR status. In HER2-negative/HR-positive group, there were 53

patients with no microcalcification, 18 with low-to-intermediate, and 12 with

highly suspicious microcalcifications. There were 478 DEGs (292 up and 186

down) in highly suspicious/no microcalcification comparison set, 236 DEGs

(182 up and 54 down) in highly suspicious/low-to-intermediate comparison

set, and 252 DEGs (103 up and 149 down) in low-to-intermediate/no

microcalcification comparison set. There was no statistically significant GO

term. However, using 478 DEGs in highly suspicious/no microcalcification

comparison set, we found that the 10 top-ranked biological functions (P value

from 10-4 to 10-2) included GO terms related with immune system when it was

ordered by P value (Figure 7).

Page 31: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

24

Figure 7. Enriched GO terms in HER2-negative/HR-positive subgroup

Page 32: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

25

In HER2-negative/HR-negative subgroup, there were 34 patients with no

microcalcification, 9 with low-to-intermediate, and 4 with highly suspicious

microcalcifications. There were 715 DEGs (310 up and 405 down) in highly

suspicious/no microcalcification comparison set, 309 DEGs (203 up and 106

down) in highly suspicious/low-to-intermediate comparison set, and 316

DEGs (159 up and 157 down) in low-to-intermediate/no microcalcification

comparison set. Using 715 DEGs in highly suspicious/no microcalcification

comparison set, we found that the 10 top-ranked biological functions (P value

from 10-28 to 10-10) included GO terms related with immune system when it

was ordered by P value (Figure 8).

Page 33: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

26

Figure 8. Enriched GO terms in HER2-negative/HR-negative subgroup

Page 34: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

27

Discussion

We have searched for DEGs in breast cancer containing

microcalcifications on mammography in compare with those without

microcalcifications. GO and clustering analysis using DAVID and BioLattice

identified genes and corresponding molecular networks that were highly

associated with the decreased immune system activity in patients with

mammographic microcalcifications. Interestingly, there were larger numbers

of DEGs according to the status of mammographic microcalcifications in

HER2-negative group than HER2-positive group. In addition, there were

many statistically significant annotation clusters composed of immune system

or inflammation related GO terms in HER2-negative group, as is not in

HER2-positive group.

These results suggest that microcalcifications of breast cancer are in

a part interact with immune system. Recently, there has been growing

interests in tumor-immune interaction; that immunity, and its mutual

interaction with tumor cells and tumor microenvironment, might play an

important role in malignancy. Moreover, there are evidences that tumor

infiltrating lymphocytes (TILs) may have prognostic value and sometimes

even predictive value in triple-negative breast cancer (32-35) or HER2-

positive breast cancer (33, 36). In other words, the presence of high TILs is

associated with a better prognosis. Relationship between tumor and immune

system is not solely an issue of immune response to tumor cells, but it also

modulates tumor microenvironment and affects the disease biology including

processes of tumor growth, metastasis, and therapeutic responsiveness (37).

Page 35: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

28

Up to date, the relationship between immune system activity and

prognosis is not clearly discovered yet in other subtypes except triple-negative

and HER2-positive breast cancer (37). Besides, there is no data whether

immune system activity has association with mammographic

microcalcifications. Even though the direct cellular mechanisms or biologic

pathways between microcalcifications and immune system are not discovered,

we can suggest one possible explanation for it. It is known that rapidly

proliferating tumor cells which use up the bloody supply result in tumor

necrosis and subsequent acidosis in the microenvironment that finally causes

calcium accumulation in the ducts (38). Activated immune system may deter

the proliferation of tumor cells and necrosis caused by hypoxia seldom occurs.

In contrary, breast cancers with decreased immune system activity likely to

have uncontrolled tumor cell proliferation and tumor necrosis may frequently

occur and it finally causes microcalcifications in the ducts. Thus, it makes

sense that breast cancers without microcalcifications are associated with

activated immune system and breast cancer with microcalcifications are

associated with decreased immune system activity. However, there is a

conflict between our results that breast cancer with microcalcifications is

associated with rapidly proliferating tumor cells following decreased immune

system and the fact that DCIS is known as indolent non-invasive tumor but

frequently associated with mammographic microcalcifications. The way how

microcalcifications are produced is might be various between invasive cancer

and DCIS and it needs to be checked in further study.

Our results indicate that there are biological differences regarding

Page 36: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

29

immune system activity between breast cancers according to the state of

microcalcifications. Although P values obtained in several subgroup analyses

didn’t reach the statistical significance probably due to small number of cases,

our results showed that the association between immune system and

microcalcifications is consistently seen in multistep subgroup analyses. If

increased activity of immune system is associated with good prognosis in

breast cancer, breast cancer with mammographic microcalcifications are

possibly associated with poor prognosis. Similarly, in HR-positive breast

cancers, absence of microcalcifications in the mass is known to be an

independent predictors associated with low recurrence score which was

assessed by the 21-gene recurrence score assays. There remains the curiosity

that why the differences is more prominent in HER2-negative subgroup than

HER2-positive subgroup. However, the recognition of this association

between mammographic microcalcifications and immune system could lead to

a better and deep understanding of the underlying tumor biology, may be

useful in dissecting complex tumor-immune interaction and in providing

potential targets for novel therapeutic approaches to breast cancer therapy

with drugs, help to identify noninvasive imaging surrogates for breast cancer,

and offer a potential strategy for noninvasively selecting patients who may be

candidates for individualized therapies (37, 39).

Our study has several limitations. First, this study is retrospective

design, and there may be selection bias in our database. In addition, we don’t

have validation set to support our results. In The Cancer Genome Atlas

(TCGA), there were only 4 patients with preoperative mammography at the

Page 37: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

30

time of our study. Second, interpretation of mammographic

microcalcifications is subjective and other imaging findings were not

considered in this study. Third, breast cancer is a heterogeneous disease

regarding its gene expression profiles. Even though we performed subgroup

analysis according to IHC-based subtypes, these are not exactly same with

molecular subtypes. Thus, there are possibilities that the differences between

two groups are not only due to microcalcifications status but also due to

intrinsic subtypes of breast cancer. Also the interaction between breast cancer

and the immune system may be different and rely on different mechanisms by

molecular subtypes or imaging phenotypes. Lastly, we didn’t perform in vitro

experiment regarding cellular mechanism of calcium deposit or ex vivo

experiment whether immune cells differently exist or immune cell markers are

differently presented according to mammographic microcalcifications indeed.

It is beyond our reach, but our results could offer a hint for further study about

biological process or cell signaling pathways of microcalcifications formation.

However, this is the first study examining the global gene expression profiles

of breast cancer according to the mammographic microcalcifications. And it is

the largest study where over one hundred patients with microarray data were

enrolled and correlated with imaging feature in breast cancer. In addition, the

results of this study indicate that molecular profiles of breast cancer with

microcalcifications are affected by HER2 status. It suggests that there might

be association (either activation or inhibition) between HER2 signaling

pathway and immune pathway in breast cancer and also proposes potentially

targetable HER2-independent and -dependent mechanisms for immune

Page 38: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

31

system in breast cancer with microcalcifications.

In conclusion, gene expression patterns are different according to the

status of mammographic microcalcifications in breast cancer. Breast cancers

with mammographic microcalcifications are associated with decreased

immune system activity. Further studies are needed to investigate the

relationship between the mammographic microcalcifications and clinical

outcome in breast cancer patients.

Page 39: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

32

References

1. Nystrom L, Andersson I, Bjurstam N, Frisell J, Nordenskjold B, Rutqvist

LE. Long-term effects of mammography screening: updated overview of the Swedish

randomised trials. Lancet. 2002;359(9310):909-19.

2. Tabar L, Tony Chen HH, Amy Yen MF, Tot T, Tung TH, Chen LS, et al.

Mammographic tumor features can predict long-term outcomes reliably in women

with 1-14-mm invasive breast carcinoma. Cancer. 2004;101(8):1745-59.

3. Tabár L, Tot T, Dean PB. Crushed stone-like calcifications: the most

frequent malignant type: Thieme; 2008.

4. Tabar L, Tot T, Dean P. Breast cancer-early detection with mammography:

Crushed stone-like calcifications: The most frequent malignant type. 2008.

5. Palka I, Ormandi K, Gaal S, Boda K, Kahan Z. Casting-type calcifications

on the mammogram suggest a higher probability of early relapse and death among

high-risk breast cancer patients. Acta oncologica. 2007;46(8):1178-83.

6. Bent CK, Bassett LW, D'Orsi CJ, Sayre JW. The positive predictive value of

BI-RADS microcalcification descriptors and final assessment categories. AJR

American journal of roentgenology. 2010;194(5):1378-83.

7. Tabar L, Chen HH, Duffy SW, Yen MF, Chiang CF, Dean PB, et al. A novel

method for prediction of long-term outcome of women with T1a, T1b, and 10-14 mm

invasive breast cancers: a prospective study. Lancet. 2000;355(9202):429-33.

8. Holmberg L, Wong YN, Tabar L, Ringberg A, Karlsson P, Arnesson LG, et

al. Mammography casting-type calcification and risk of local recurrence in DCIS:

analyses from a randomised study. British journal of cancer. 2013;108(4):812-9.

9. Ling H, Liu ZB, Xu LH, Xu XL, Liu GY, Shao ZM. Malignant calcification

is an important unfavorable prognostic factor in primary invasive breast cancer. Asia-

Pacific journal of clinical oncology. 2013;9(2):139-45.

Page 40: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

33

10. Bennett RL, Evans AJ, Kutt E, Record C, Bobrow LG, Ellis IO, et al.

Pathological and mammographic prognostic factors for screen detected cancers in a

multi-centre randomised, controlled trial of mammographic screening in women from

age 40 to 48 years. Breast. 2011;20(6):525-8.

11. Gajdos C, Tartter PI, Bleiweiss IJ, Hermann G, de Csepel J, Estabrook A, et

al. Mammographic appearance of nonpalpable breast cancer reflects pathologic

characteristics. Annals of surgery. 2002;235(2):246-51.

12. Evans AJ, Pinder SE, Ellis IO, Sibbering DM, Elston CW, Poller DN, et al.

Correlations between the mammographic features of ductal carcinoma in situ (DCIS)

and C-erbB-2 oncogene expression. Nottingham Breast Team. Clin Radiol.

1994;49(8):559-62.

13. Seo BK, Pisano ED, Kuzimak CM, Koomen M, Pavic D, Lee Y, et al.

Correlation of HER-2/neu overexpression with mammography and age distribution in

primary breast carcinomas. Academic radiology. 2006;13(10):1211-8.

14. Karamouzis MV, Likaki-Karatza E, Ravazoula P, Badra FA, Koukouras D,

Tzorakoleftherakis E, et al. Non-palpable breast carcinomas: correlation of

mammographically detected malignant-appearing microcalcifications and molecular

prognostic factors. International journal of cancer Journal international du cancer.

2002;102(1):86-90.

15. Sun SS, Zhang B, Zhao HM, Cao XC. Association between mammographic

features and clinicopathological characteristics in invasive ductal carcinoma of breast

cancer. Molecular and clinical oncology. 2014;2(4):623-9.

16. Yamamoto S, Maki DD, Korn RL, Kuo MD. Radiogenomic analysis of

breast cancer using MRI: a preliminary study to define the landscape. AJR American

journal of roentgenology. 2012;199(3):654-63.

17. Wan T, Bloch BN, Plecha D, Thompson CL, Gilmore H, Jaffe C, et al. A

Page 41: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

34

Radio-genomics Approach for Identifying High Risk Estrogen Receptor-positive

Breast Cancers on DCE-MRI: Preliminary Results in Predicting OncotypeDX Risk

Scores. Scientific reports. 2016;6:21394.

18. Zhu Y, Li H, Guo W, Drukker K, Lan L, Giger ML, et al. Deciphering

Genomic Underpinnings of Quantitative MRI-based Radiomic Phenotypes of

Invasive Breast Carcinoma. Scientific reports. 2015;5:17787.

19. Yamamoto S, Han W, Kim Y, Du L, Jamshidi N, Huang D, et al. Breast

Cancer: Radiogenomic Biomarker Reveals Associations among Dynamic Contrast-

enhanced MR Imaging, Long Noncoding RNA, and Metastasis. Radiology.

2015;275(2):384-92.

20. Wang X, Chao L, Chen L, Tian B, Ma G, Zang Y, et al. Correlation of

mammographic calcifications with Her-2/neu overexpression in primary breast

carcinomas. Journal of digital imaging. 2008;21(2):170-6.

21. Wolff AC, Hammond ME, Schwartz JN, Hagerty KL, Allred DC, Cote RJ,

et al. American Society of Clinical Oncology/College of American Pathologists

guideline recommendations for human epidermal growth factor receptor 2 testing in

breast cancer. Archives of pathology & laboratory medicine. 2007;131(1):18-43.

22. Imbeaud S, Graudens E, Boulanger V, Barlet X, Zaborski P, Eveno E, et al.

Towards standardization of RNA quality assessment using user-independent

classifiers of microcapillary electrophoresis traces. Nucleic acids research.

2005;33(6):e56.

23. Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf

U, et al. Exploration, normalization, and summaries of high density oligonucleotide

array probe level data. Biostatistics. 2003;4(2):249-64.

24. Carvalho BS, Irizarry RA. A framework for oligonucleotide microarray

preprocessing. Bioinformatics. 2010;26(19):2363-7.

Page 42: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

35

25. Gautier L, Cope L, Bolstad BM, Irizarry RA. affy--analysis of Affymetrix

GeneChip data at the probe level. Bioinformatics. 2004;20(3):307-15.

26. Kerr MK, Martin M, Churchill GA. Analysis of variance for gene

expression microarray data. Journal of computational biology : a journal of

computational molecular cell biology. 2000;7(6):819-37.

27. Huang da W, Sherman BT, Lempicki RA. Systematic and integrative

analysis of large gene lists using DAVID bioinformatics resources. Nature protocols.

2009;4(1):44-57.

28. Huang da W, Sherman BT, Lempicki RA. Bioinformatics enrichment tools:

paths toward the comprehensive functional analysis of large gene lists. Nucleic acids

research. 2009;37(1):1-13.

29. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H, Cherry JM, et al.

Gene ontology: tool for the unification of biology. The Gene Ontology Consortium.

Nature genetics. 2000;25(1):25-9.

30. Kim J, Chung HJ, Jung Y, Kim KK, Kim JH. BioLattice: a framework for

the biological interpretation of microarray gene expression data using concept lattice

analysis. Journal of biomedical informatics. 2008;41(2):232-41.

31. Tavazoie S, Hughes JD, Campbell MJ, Cho RJ, Church GM. Systematic

determination of genetic network architecture. Nature genetics. 1999;22(3):281-5.

32. Loi S, Sirtaine N, Piette F, Salgado R, Viale G, Van Eenoo F, et al.

Prognostic and predictive value of tumor-infiltrating lymphocytes in a phase III

randomized adjuvant breast cancer trial in node-positive breast cancer comparing the

addition of docetaxel to doxorubicin with doxorubicin-based chemotherapy: BIG 02-

98. Journal of clinical oncology : official journal of the American Society of Clinical

Oncology. 2013;31(7):860-7.

33. Denkert C, von Minckwitz G, Brase JC, Sinn BV, Gade S, Kronenwett R, et

Page 43: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

36

al. Tumor-infiltrating lymphocytes and response to neoadjuvant chemotherapy with or

without carboplatin in human epidermal growth factor receptor 2-positive and triple-

negative primary breast cancers. Journal of clinical oncology : official journal of the

American Society of Clinical Oncology. 2015;33(9):983-91.

34. Dieci MV, Mathieu MC, Guarneri V, Conte P, Delaloge S, Andre F, et al.

Prognostic and predictive value of tumor-infiltrating lymphocytes in two phase III

randomized adjuvant breast cancer trials. Annals of oncology : official journal of the

European Society for Medical Oncology / ESMO. 2015;26(8):1698-704.

35. Ibrahim EM, Al-Foheidi ME, Al-Mansour MM, Kazkaz GA. The prognostic

value of tumor-infiltrating lymphocytes in triple-negative breast cancer: a meta-

analysis. Breast cancer research and treatment. 2014;148(3):467-76.

36. Bianchini G, Gianni L. The immune system and response to HER2-targeted

treatment in breast cancer. The Lancet Oncology. 2014;15(2):e58-68.

37. Dieci MV, Griguolo G, Miglietta F, Guarneri V. The immune system and

hormone-receptor positive breast cancer: Is it really a dead end? Cancer treatment

reviews. 2016;46:9-19.

38. Tse GM, Tan PH, Cheung HS, Chu WC, Lam WW. Intermediate to highly

suspicious calcification in breast lesions: a radio-pathologic correlation. Breast cancer

research and treatment. 2008;110(1):1-7.

39. Diehn M, Nardini C, Wang DS, McGovern S, Jayaraman M, Liang Y, et al.

Identification of noninvasive imaging surrogates for brain tumor gene-expression

modules. Proceedings of the National Academy of Sciences of the United States of

America. 2008;105(13):5213-8.

Page 44: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

37

Appendix 1. Differentially expressed genes in highly suspicious

microcalcifications /no microcalcification comparison set in total population

(top 200 genes are shown, ordered by P value)

NCBI ID

HUGO Gene Name P value

94103 ORMDL3 ORMDL sphingolipid biosynthesis regulator 3

0.000

3038 HAS3 hyaluronan synthase 3 0.000 338557 FFAR4 free fatty acid receptor 4 0.000 64175 P3H1 prolyl 3-hydroxylase 1 0.000 1301 COL11A

1 collagen, type XI, alpha 1 0.000

54894 RNF43 ring finger protein 43 0.000 84624 FNDC1 fibronectin type III domain containing 1 0.000 6507 SLC1A3 solute carrier family 1 (glial high affinity

glutamate transporter), member 3 0.000

3983 ABLIM1 actin binding LIM protein 1 0.000 5996 RGS1 regulator of G-protein signaling 1 0.000 1501 CTNND2 catenin (cadherin-associated protein), delta 2 0.000 2064 ERBB2 erb-b2 receptor tyrosine kinase 2 0.000 80896 NPL N-acetylneuraminate pyruvate lyase

(dihydrodipicolinate synthase) 0.000

5328 PLAU plasminogen activator, urokinase 0.000 713 C1QB complement component 1, q subcomponent,

B chain 0.000

822 CAPG capping protein (actin filament), gelsolin-like

0.000

1573 CYP2J2 cytochrome P450, family 2, subfamily J, polypeptide 2

0.000

586 BCAT1 branched chain amino-acid transaminase 1, cytosolic

0.000

2706 GJB2 gap junction protein beta 2 0.000 10082 GPC6 glypican 6 0.000 8654 PDE5A phosphodiesterase 5A, cGMP-specific 0.000 284161 GDPD1 glycerophosphodiester phosphodiesterase

domain containing 1 0.000

7045 TGFBI transforming growth factor, beta-induced, 68kDa

0.000

29108 PYCARD PYD and CARD domain containing 0.000 4804 NGFR nerve growth factor receptor 0.000 1508 CTSB cathepsin B 0.000 2207 FCER1G Fc fragment of IgE, high affinity I, receptor

for; gamma polypeptide 0.000

860 RUNX2 runt-related transcription factor 2 0.000

Page 45: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

38

NCBI ID

HUGO Gene Name P value

639 PRDM1 PR domain containing 1, with ZNF domain 0.000 4973 OLR1 oxidized low density lipoprotein (lectin-like)

receptor 1 0.000

5329 PLAUR plasminogen activator, urokinase receptor 0.000 113451 AZIN2 antizyme inhibitor 2 0.000 6578 SLCO2A

1 solute carrier organic anion transporter family, member 2A1

0.000

2886 GRB7 growth factor receptor-bound protein 7 0.000 219 ALDH1B

1 aldehyde dehydrogenase 1 family, member B1

0.000

7805 LAPTM5 lysosomal protein transmembrane 5 0.000 79783 SUGCT succinyl-CoA:glutarate-CoA transferase 0.000 2532 ACKR1 atypical chemokine receptor 1 (Duffy blood

group) 0.000

27244 SESN1 sestrin 1 0.000 79899 PRR5L proline rich 5 like 0.000 84299 MIEN1 migration and invasion enhancer 1 0.000 2205 FCER1A Fc fragment of IgE, high affinity I, receptor

for; alpha polypeptide 0.000

8828 NRP2 neuropilin 2 0.000 91614 DEPDC7 DEP domain containing 7 0.000 90865 IL33 interleukin 33 0.000 54407 SLC38A2 solute carrier family 38, member 2 0.000 2335 FN1 fibronectin 1 0.000 10797 MTHFD2 methylenetetrahydrofolate dehydrogenase

(NADP+ dependent) 2, methenyltetrahydrofolate cyclohydrolase

0.000

2517 FUCA1 fucosidase, alpha-L- 1, tissue 0.000 399664 MEX3D mex-3 RNA binding family member D 0.000 339745 SPOPL speckle-type POZ protein-like 0.000 79762 C1orf115 chromosome 1 open reading frame 115 0.000 55080 TAPBPL TAP binding protein-like 0.000 55876 GSDMB gasdermin B 0.000 440021 KRTAP5-

2 keratin associated protein 5-2 0.000

154661 RUNDC3B

RUN domain containing 3B 0.000

1244 ABCC2 ATP-binding cassette, sub-family C (CFTR/MRP), member 2

0.000

5580 PRKCD protein kinase C, delta 0.000 10437 IFI30 interferon, gamma-inducible protein 30 0.000 345274 SLC10A6 solute carrier family 10 (sodium/bile acid

cotransporter), member 6 0.000

11031 RAB31 RAB31, member RAS oncogene family 0.000

Page 46: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

39

NCBI ID

HUGO Gene Name P value

100506144

ZMYM6NB

ZMYM6 neighbor 0.000

83699 SH3BGRL2

SH3 domain binding glutamate-rich protein like 2

0.000

23384 SPECC1L sperm antigen with calponin homology and coiled-coil domains 1-like

0.000

79983 POF1B premature ovarian failure, 1B 0.000 1215 CMA1 chymase 1, mast cell 0.000 9635 CLCA2 chloride channel accessory 2 0.000 51705 EMCN endomucin 0.000 93210 PGAP3 post-GPI attachment to proteins 3 0.000 57609 DIP2B disco-interacting protein 2 homolog B 0.000 3652 IPP intracisternal A particle-promoted

polypeptide 0.000

92 ACVR2A activin A receptor type IIA 0.000 199 AIF1 allograft inflammatory factor 1 0.000 968 CD68 CD68 molecule 0.000 220441 RNF152 ring finger protein 152 0.000 9914 ATP2C2 ATPase, Ca++ transporting, type 2C,

member 2 0.000

3156 HMGCR 3-hydroxy-3-methylglutaryl-CoA reductase 0.000 729220 FLJ45513 uncharacterized LOC729220 0.000 23411 SIRT1 sirtuin 1 0.000 131578 LRRC15 leucine rich repeat containing 15 0.000 8590 OR6A2 olfactory receptor, family 6, subfamily A,

member 2 0.000

3021 H3F3B H3 histone, family 3B (H3.3B) 0.000 5138 PDE2A phosphodiesterase 2A, cGMP-stimulated 0.000 254170 FBXO33 F-box protein 33 0.000 54532 USP53 ubiquitin specific peptidase 53 0.000 2882 GPX7 glutathione peroxidase 7 0.000 84181 CHD6 chromodomain helicase DNA binding

protein 6 0.000

26064 RAI14 retinoic acid induced 14 0.000 23598 PATZ1 POZ (BTB) and AT hook containing zinc

finger 1 0.000

5169 ENPP3 ectonucleotide pyrophosphatase/phosphodiesterase 3

0.000

4481 MSR1 macrophage scavenger receptor 1 0.000 9618 TRAF4 TNF receptor-associated factor 4 0.000 8452 CUL3 cullin 3 0.000 51652 CHMP3 charged multivesicular body protein 3 0.000 2212 FCGR2A Fc fragment of IgG, low affinity IIa, receptor

(CD32) 0.000

Page 47: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

40

NCBI ID

HUGO Gene Name P value

2590 GALNT2 polypeptide N-acetylgalactosaminyltransferase 2

0.000

4323 MMP14 matrix metallopeptidase 14 (membrane-inserted)

0.000

59351 PBOV1 prostate and breast cancer overexpressed 1 0.000 50515 CHST11 carbohydrate (chondroitin 4)

sulfotransferase 11 0.000

114784 CSMD2 CUB and Sushi multiple domains 2 0.000 57194 ATP10A ATPase, class V, type 10A 0.000 84640 USP38 ubiquitin specific peptidase 38 0.000 23505 TMEM13

1 transmembrane protein 131 0.000

151473 SLC16A14

solute carrier family 16, member 14 0.000

124801 LSM12 LSM12 homolog 0.000 115908 CTHRC1 collagen triple helix repeat containing 1 0.000 7226 TRPM2 transient receptor potential cation channel,

subfamily M, member 2 0.000

5789 PTPRD protein tyrosine phosphatase, receptor type, D

0.000

5502 PPP1R1A protein phosphatase 1, regulatory (inhibitor) subunit 1A

0.000

2745 GLRX glutaredoxin (thioltransferase) 0.000 81832 NETO1 neuropilin (NRP) and tolloid (TLL)-like 1 0.000 6567 SLC16A2 solute carrier family 16, member 2 (thyroid

hormone transporter) 0.000

84311 MRPL45 mitochondrial ribosomal protein L45 0.000 1182 CLCN3 chloride channel, voltage-sensitive 3 0.000 10855 HPSE heparanase 0.000 196051 PLPP4 phospholipid phosphatase 4 0.000 54209 TREM2 triggering receptor expressed on myeloid

cells 2 0.000

2210 FCGR1B Fc fragment of IgG, high affinity Ib, receptor (CD64)

0.000

25903 OLFML2B

olfactomedin like 2B 0.000

6426 SRSF1 serine/arginine-rich splicing factor 1 0.000 7130 TNFAIP6 tumor necrosis factor, alpha-induced protein

6 0.000

2825 GPR1 G protein-coupled receptor 1 0.000 5709 PSMD3 proteasome 26S subunit, non-ATPase 3 0.000 55083 KIF26B kinesin family member 26B 0.000 220929 ZNF438 zinc finger protein 438 0.000 30837 SOCS7 suppressor of cytokine signaling 7 0.000

Page 48: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

41

NCBI ID

HUGO Gene Name P value

55808 ST6GALNAC1

ST6 (alpha-N-acetyl-neuraminyl-2,3-beta-galactosyl-1,3)-N-acetylgalactosaminide alpha-2,6-sialyltransferase 1

0.000

719 C3AR1 complement component 3a receptor 1 0.000 4322 MMP13 matrix metallopeptidase 13 0.000 23601 CLEC5A C-type lectin domain family 5, member A 0.000 3682 ITGAE integrin, alpha E (antigen CD103, human

mucosal lymphocyte antigen 1; alpha polypeptide)

0.000

79953 SYNDIG1

synapse differentiation inducing 1 0.000

3107 HLA-C major histocompatibility complex, class I, C 0.000 7515 XRCC1 X-ray repair complementing defective repair

in Chinese hamster cells 1 0.000

1950 EGF epidermal growth factor 0.000 168507 PKD1L1 polycystic kidney disease 1 like 1 0.000 7783 ZP2 zona pellucida glycoprotein 2 (sperm

receptor) 0.000

401494 HACD4 3-hydroxyacyl-CoA dehydratase 4 0.000 9659 PDE4DIP phosphodiesterase 4D interacting protein 0.000 3081 HGD homogentisate 1,2-dioxygenase 0.000 619189 SERINC4 serine incorporator 4 0.000 5480 PPIC peptidylprolyl isomerase C (cyclophilin C) 0.000 8832 CD84 CD84 molecule 0.000 283209 PGM2L1 phosphoglucomutase 2-like 1 0.000 9411 ARHGAP

29 Rho GTPase activating protein 29 0.000

51300 TIMMDC1

translocase of inner mitochondrial membrane domain containing 1

0.000

54972 TMEM132A

transmembrane protein 132A 0.000

8897 MTMR3 myotubularin related protein 3 0.000 339479 BRINP3 bone morphogenetic protein/retinoic acid

inducible neural-specific 3 0.000

56944 OLFML3 olfactomedin like 3 0.000 221895 JAZF1 JAZF zinc finger 1 0.000 4354 MPP1 membrane protein, palmitoylated 1 0.000 1363 CPE carboxypeptidase E 0.000 5045 FURIN furin (paired basic amino acid cleaving

enzyme) 0.000

220002 CYB561A3

cytochrome b561 family, member A3 0.000

11076 TPPP tubulin polymerization promoting protein 0.000 57214 CEMIP cell migration inducing protein, hyaluronan

binding 0.000

Page 49: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

42

NCBI ID

HUGO Gene Name P value

3763 KCNJ6 potassium channel, inwardly rectifying subfamily J, member 6

0.000

911 CD1C CD1c molecule 0.000 6404 SELPLG selectin P ligand 0.000 6558 SLC12A2 solute carrier family 12

(sodium/potassium/chloride transporter), member 2

0.000

421 ARVCF armadillo repeat gene deleted in velocardiofacial syndrome

0.000

730 C7 complement component 7 0.000 7148 TNXB tenascin XB 0.000 285180 RUFY4 RUN and FYVE domain containing 4 0.000 27036 SIGLEC7 sialic acid binding Ig-like lectin 7 0.000 92745 SLC38A5 solute carrier family 38, member 5 0.000 200734 SPRED2 sprouty-related, EVH1 domain containing 2 0.001 79720 VPS37B vacuolar protein sorting 37 homolog B (S.

cerevisiae) 0.001

1848 DUSP6 dual specificity phosphatase 6 0.001 91 ACVR1B activin A receptor type IB 0.001 85379 KIAA167

1 KIAA1671 0.001

5787 PTPRB protein tyrosine phosphatase, receptor type, B

0.001

5027 P2RX7 purinergic receptor P2X, ligand gated ion channel, 7

0.001

1741 DLG3 discs, large homolog 3 (Drosophila) 0.001 80045 GPR157 G protein-coupled receptor 157 0.001 256979 SUN3 Sad1 and UNC84 domain containing 3 0.001 10154 PLXNC1 plexin C1 0.001 54708 MARCH5 membrane-associated ring finger (C3HC4) 5 0.001 5199 CFP complement factor properdin 0.001 8030 CCDC6 coiled-coil domain containing 6 0.001 317649 EIF4E3 eukaryotic translation initiation factor 4E

family member 3 0.001

1303 COL12A1

collagen, type XII, alpha 1 0.001

11094 CACFD1 calcium channel flower domain containing 1 0.001 56889 TM9SF3 transmembrane 9 superfamily member 3 0.001 168391 GALNTL

5 polypeptide N-acetylgalactosaminyltransferase-like 5

0.001

1833 EPYC epiphycan 0.001 121457 IKBIP IKBKB interacting protein 0.001 120224 TMEM45

B transmembrane protein 45B 0.001

Page 50: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

43

NCBI ID

HUGO Gene Name P value

10039 PARP3 poly(ADP-ribose) polymerase family member 3

0.001

89790 SIGLEC10

sialic acid binding Ig-like lectin 10 0.001

387264 KRTAP5-1

keratin associated protein 5-1 0.001

5783 PTPN13 protein tyrosine phosphatase, non-receptor type 13 (APO-1/CD95 (Fas)-associated phosphatase)

0.001

526 ATP6V1B2

ATPase, H+ transporting, lysosomal 56/58kDa, V1 subunit B2

0.001

338657 CCDC84 coiled-coil domain containing 84 0.001 10948 STARD3 StAR-related lipid transfer domain

containing 3 0.001

441272 SPDYE3 speedy/RINGO cell cycle regulator family member E3

0.001

118424 UBE2J2 ubiquitin-conjugating enzyme E2, J2 0.001 8710 SERPINB

7 serpin peptidase inhibitor, clade B (ovalbumin), member 7

0.001

692094 MSMP microseminoprotein, prostate associated 0.001

Page 51: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

44

Appendix 2. Functional annotation clusters using DEGs (n = 1838) from highly suspicious/no microcalcifications comparison set which

contain GO terms in the biological process category with enrichment score > 3.5

Gene Ontology ID Term Fold Enrichment

P Value Bonferroni

Annotation Cluster 1 Enrichment Score: 8.451 GO:0006952 defense response 2.362 0.000 0.000 GO:0009611 response to wounding 2.291 0.000 0.000 GO:0006954 inflammatory response 2.668 0.000 0.000 Annotation Cluster 2 Enrichment Score: 7.474 GO:0019882 antigen processing and presentation 6.530 0.000 0.000 GO:0048002 antigen processing and presentation of peptide antigen 11.614 0.000 0.000 GO:0019884 antigen processing and presentation of exogenous antigen 13.937 0.000 0.000 GO:0002478 antigen processing and presentation of exogenous peptide

antigen 15.767 0.000 0.000

GO:0002474 antigen processing and presentation of peptide antigen via MHC class I

11.477 0.000 0.001

GO:0019886 antigen processing and presentation of exogenous peptide antigen via MHC class II

16.260 0.000 0.029

GO:0002495 antigen processing and presentation of peptide antigen via MHC class II

16.260 0.000 0.029

GO:0002504 antigen processing and presentation of peptide or polysaccharide antigen via MHC class II

6.570 0.000 0.036

Annotation Cluster 3 Enrichment Score: 5.300 GO:0002684 positive regulation of immune system process 3.370 0.000 0.000

Page 52: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

45

Gene Ontology ID Term Fold Enrichment

P Value Bonferroni

GO:0050778 positive regulation of immune response 4.037 0.000 0.000 GO:0048584 positive regulation of response to stimulus 2.940 0.000 0.000 GO:0002252 immune effector process 3.559 0.000 0.002 GO:0002443 leukocyte mediated immunity 4.285 0.000 0.005 GO:0002250 adaptive immune response 4.505 0.000 0.005 GO:0002460 adaptive immune response based on somatic recombination

of immune receptors built from immunoglobulin superfamily domains

4.505 0.000 0.005

GO:0002449 lymphocyte mediated immunity 4.336 0.000 0.044 GO:0016064 immunoglobulin mediated immune response 4.818 0.000 0.079 GO:0019724 B cell mediated immunity 4.646 0.000 0.111 GO:0002253 activation of immune response 3.459 0.000 0.242 GO:0006958 complement activation, classical pathway 3.738 0.042 1.000 GO:0002455 humoral immune response mediated by circulating

immunoglobulin 3.497 0.052 1.000

Annotation Cluster 4 Enrichment Score: 4.538 GO:0032963 collagen metabolic process 7.743 0.000 0.008 GO:0044236 multicellular organismal metabolic process 6.445 0.000 0.013 GO:0044259 multicellular organismal macromolecule metabolic process 6.993 0.000 0.020 GO:0044243 multicellular organismal catabolic process 6.671 0.000 0.313 GO:0030574 collagen catabolic process 6.504 0.002 0.994 Annotation Cluster 5 Enrichment Score: 4.124 GO:0030199 collagen fibril organization 7.476 0.000 0.011 GO:0030198 extracellular matrix organization 3.335 0.000 0.206 GO:0043062 extracellular structure organization 2.394 0.001 0.982

Page 53: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

46

Gene Ontology ID Term Fold Enrichment

P Value Bonferroni

Annotation Cluster 6 Enrichment Score: 3.531 GO:0001817 regulation of cytokine production 3.833 0.000 0.000 GO:0001819 positive regulation of cytokine production 4.095 0.000 0.009 GO:0050707 regulation of cytokine secretion 7.226 0.000 0.015 GO:0050714 positive regulation of protein secretion 5.816 0.000 0.035 GO:0051240 positive regulation of multicellular organismal process 2.488 0.000 0.062 GO:0051222 positive regulation of protein transport 4.206 0.000 0.129 GO:0050715 positive regulation of cytokine secretion 7.541 0.000 0.146 GO:0050708 regulation of protein secretion 4.485 0.000 0.152 GO:0032731 positive regulation of interleukin-1 beta production 9.291 0.000 0.565 GO:0051223 regulation of protein transport 2.853 0.001 0.882 GO:0032732 positive regulation of interleukin-1 production 7.652 0.001 0.899 GO:0070201 regulation of establishment of protein localization 2.688 0.001 0.979 GO:0050718 positive regulation of interleukin-1 beta secretion 9.033 0.002 0.991 GO:0032651 regulation of interleukin-1 beta production 6.504 0.002 0.994 GO:0050706 regulation of interleukin-1 beta secretion 8.338 0.002 0.999 GO:0050716 positive regulation of interleukin-1 secretion 7.743 0.003 1.000 GO:0032652 regulation of interleukin-1 production 5.656 0.003 1.000 GO:0050704 regulation of interleukin-1 secretion 7.226 0.004 1.000 GO:0051047 positive regulation of secretion 2.586 0.004 1.000 GO:0032880 regulation of protein localization 2.356 0.005 1.000 GO:0051046 regulation of secretion 1.932 0.012 1.000 GO:0051050 positive regulation of transport 1.847 0.015 1.000 GO:0060341 regulation of cellular localization 1.748 0.021 1.000

Page 54: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

47

Appendix 3. Gene lists (n = 279) which correlate with HER2 score (0, 1+, 2+,

3+) with Bonferroni corrected P value < 0.05 (ordered by P value and

correlation coefficient (r))

NCBI ID

HUGO Gene Symbol

Gene Name r

2064 ERBB2 erb-b2 receptor tyrosine kinase 2 0.716 2886 GRB7 growth factor receptor-bound protein 7 0.652 10948 STARD

3 StAR-related lipid transfer domain containing 3

0.633

84299 MIEN1 migration and invasion enhancer 1 0.603 93210 PGAP3 post-GPI attachment to proteins 3 0.676 389816 LRRC2

6 leucine rich repeat containing 26 0.591

23199 GSE1 Gse1 coiled-coil protein 0.555 118430 MUCL1 mucin-like 1 0.536 646424 SPINK8 serine peptidase inhibitor, Kazal type 8

(putative) 0.522

8557 TCAP titin-cap 0.515 51755 CDK12 cyclin-dependent kinase 12 0.512 9635 CLCA2 chloride channel accessory 2 0.511 120224 TMEM4

5B transmembrane protein 45B 0.510

25803 SPDEF SAM pointed domain containing ETS transcription factor

0.509

80021 TMEM62

transmembrane protein 62 0.509

144110 TMEM86A

transmembrane protein 86A 0.502

94103 ORMDL3

ORMDL sphingolipid biosynthesis regulator 3

0.500

5409 PNMT phenylethanolamine N-methyltransferase 0.498 3081 HGD homogentisate 1,2-dioxygenase 0.498 2863 GPR39 G protein-coupled receptor 39 0.497 27289 RND1 Rho family GTPase 1 0.495 80725 SRCIN1 SRC kinase signaling inhibitor 1 0.489 7703 PCGF2 polycomb group ring finger 2 0.482 1741 DLG3 discs, large homolog 3 (Drosophila) 0.481 26154 ABCA1

2 ATP-binding cassette, sub-family A (ABC1), member 12

0.480

412 STS steroid sulfatase (microsomal), isozyme S 0.480 145741 C2CD4

A C2 calcium-dependent domain containing 4A

0.478

Page 55: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

48

NCBI ID

HUGO Gene Symbol

Gene Name r

2108 ETFA electron-transfer-flavoprotein, alpha polypeptide

0.478

8659 ALDH4A1

aldehyde dehydrogenase 4 family, member A1

0.478

79017 GGCT gamma-glutamylcyclotransferase 0.476 3110 MNX1 motor neuron and pancreas homeobox 1 0.475 92291 CAPN1

3 calpain 13 0.470

2539 G6PD glucose-6-phosphate dehydrogenase 0.470 5083 PAX9 paired box 9 0.466 5514 PPP1R1

0 protein phosphatase 1, regulatory subunit 10

0.464

3227 HOXC11

homeobox C11 0.461

57415 C3orf14 chromosome 3 open reading frame 14 0.460 197131 UBR1 ubiquitin protein ligase E3 component n-

recognin 1 0.460

56241 SUSD2 sushi domain containing 2 0.459 2264 FGFR4 fibroblast growth factor receptor 4 0.453 5469 MED1 mediator complex subunit 1 0.453 151473 SLC16

A14 solute carrier family 16, member 14 0.453

283987 HID1 HID1 domain containing 0.451 339287 MSL1 male-specific lethal 1 homolog

(Drosophila) 0.451

6567 SLC16A2

solute carrier family 16, member 2 (thyroid hormone transporter)

0.449

389434 IYD iodotyrosine deiodinase 0.448 79820 CATSP

ERB catsper channel auxiliary subunit beta 0.448

2624 GATA2 GATA binding protein 2 0.445 11094 CACFD

1 calcium channel flower domain containing 1

0.444

64090 GAL3ST2

galactose-3-O-sulfotransferase 2 0.444

57222 ERGIC1 endoplasmic reticulum-golgi intermediate compartment 1

0.443

8396 PIP4K2B

phosphatidylinositol-5-phosphate 4-kinase, type II, beta

0.443

57221 ARFGEF3

ARFGEF family member 3 0.441

29095 ORMDL2

ORMDL sphingolipid biosynthesis regulator 2

0.440

Page 56: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

49

NCBI ID

HUGO Gene Symbol

Gene Name r

200734 SPRED2

sprouty-related, EVH1 domain containing 2 0.440

157638 FAM84B

family with sequence similarity 84, member B

0.440

1244 ABCC2 ATP-binding cassette, sub-family C (CFTR/MRP), member 2

0.439

3263 HPX hemopexin 0.439 2257 FGF12 fibroblast growth factor 12 0.438 950 SCARB

2 scavenger receptor class B, member 2 0.434

284106 CISD3 CDGSH iron sulfur domain 3 0.433 7021 TFAP2

B transcription factor AP-2 beta (activating enhancer binding protein 2 beta)

0.432

5571 PRKAG1

protein kinase, AMP-activated, gamma 1 non-catalytic subunit

0.432

55890 GPRC5C

G protein-coupled receptor, class C, group 5, member C

0.432

51474 LIMA1 LIM domain and actin binding 1 0.430 9919 SEC16

A SEC16 homolog A, endoplasmic reticulum export factor

0.430

388886 LRRC75B

leucine rich repeat containing 75B 0.430

4784 NFIX nuclear factor I/X (CCAAT-binding transcription factor)

-0.428

85320 ABCC11

ATP-binding cassette, sub-family C (CFTR/MRP), member 11

0.428

57156 TMEM63C

transmembrane protein 63C 0.428

79152 FA2H fatty acid 2-hydroxylase 0.427 92840 REEP6 receptor accessory protein 6 0.427 79983 POF1B premature ovarian failure, 1B 0.427 79919 C2orf54 chromosome 2 open reading frame 54 0.427 9865 TRIL TLR4 interactor with leucine-rich repeats 0.426 2877 GPX2 glutathione peroxidase 2 0.426 399664 MEX3D mex-3 RNA binding family member D 0.424 27324 TOX3 TOX high mobility group box family

member 3 0.423

1728 NQO1 NAD(P)H dehydrogenase, quinone 1 0.423 84152 PPP1R1

B protein phosphatase 1, regulatory (inhibitor) subunit 1B

0.423

54848 ARHGEF38

Rho guanine nucleotide exchange factor 38 0.423

Page 57: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

50

NCBI ID

HUGO Gene Symbol

Gene Name r

4241 MFI2 antigen p97 (melanoma associated) identified by monoclonal antibodies 133.2 and 96.5

-0.421

29842 TFCP2L1

transcription factor CP2-like 1 -0.419

222 ALDH3B2

aldehyde dehydrogenase 3 family, member B2

0.419

388125 C2CD4B

C2 calcium-dependent domain containing 4B

0.419

79666 PLEKHF2

pleckstrin homology domain containing, family F (with FYVE domain) member 2

0.419

388633 LDLRAD1

low density lipoprotein receptor class A domain containing 1

0.419

10612 TRIM3 tripartite motif containing 3 0.418 5874 RAB27

B RAB27B, member RAS oncogene family 0.418

6337 SCNN1A

sodium channel, non voltage gated 1 alpha subunit

0.418

283209 PGM2L1

phosphoglucomutase 2-like 1 0.417

3169 FOXA1 forkhead box A1 0.415 3156 HMGC

R 3-hydroxy-3-methylglutaryl-CoA reductase 0.415

8848 TSC22D1

TSC22 domain family, member 1 0.414

79710 MORC4 MORC family CW-type zinc finger 4 0.414 79962 DNAJC

22 DnaJ (Hsp40) homolog, subfamily C, member 22

0.412

10473 HMGN4

high mobility group nucleosomal binding domain 4

-0.412

5709 PSMD3 proteasome 26S subunit, non-ATPase 3 0.411 10126 DNAL4 dynein, axonemal, light chain 4 0.410 3185 HNRNP

F heterogeneous nuclear ribonucleoprotein F 0.410

23329 TBC1D30

TBC1 domain family, member 30 0.410

6549 SLC9A2

solute carrier family 9, subfamily A (NHE2, cation proton antiporter 2), member 2

0.409

57136 APMAP adipocyte plasma membrane associated protein

0.409

2525 FUT3 fucosyltransferase 3 (galactoside 3(4)-L-fucosyltransferase, Lewis blood group)

0.409

9545 RAB3D RAB3D, member RAS oncogene family 0.409

Page 58: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

51

NCBI ID

HUGO Gene Symbol

Gene Name r

85014 TMEM141

transmembrane protein 141 0.409

1397 CRIP2 cysteine-rich protein 2 0.407 79639 TMEM5

3 transmembrane protein 53 0.405

1848 DUSP6 dual specificity phosphatase 6 0.404 5691 PSMB3 proteasome subunit beta 3 0.404 3226 HOXC1

0 homeobox C10 0.404

2582 GALE UDP-galactose-4-epimerase 0.404 100128927

ZBTB42

zinc finger and BTB domain containing 42 0.404

6307 MSMO1

methylsterol monooxygenase 1 0.403

257019 FRMD3 FERM domain containing 3 -0.403 22845 DOLK dolichol kinase 0.402 3927 LASP1 LIM and SH3 protein 1 0.402 90007 MIDN midnolin 0.402 4680 CEACA

M6 carcinoembryonic antigen-related cell adhesion molecule 6 (non-specific cross reacting antigen)

0.401

81031 SLC2A10

solute carrier family 2 (facilitated glucose transporter), member 10

0.401

80320 SP6 Sp6 transcription factor 0.400 84879 MFSD2

A major facilitator superfamily domain containing 2A

0.400

65084 TMEM135

transmembrane protein 135 0.398

3694 ITGB6 integrin, beta 6 0.398 85301 COL27

A1 collagen, type XXVII, alpha 1 -0.398

58476 TP53INP2

tumor protein p53 inducible nuclear protein 2

0.397

284161 GDPD1 glycerophosphodiester phosphodiesterase domain containing 1

0.397

326625 MMAB methylmalonic aciduria (cobalamin deficiency) cblB type

0.396

83999 KREMEN1

kringle containing transmembrane protein 1 0.396

84961 FBXL20

F-box and leucine-rich repeat protein 20 0.396

79762 C1orf115

chromosome 1 open reading frame 115 0.394

Page 59: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

52

NCBI ID

HUGO Gene Symbol

Gene Name r

83930 STARD3NL

STARD3 N-terminal like 0.394

92714 ARRDC1

arrestin domain containing 1 0.393

8564 KMO kynurenine 3-monooxygenase (kynurenine 3-hydroxylase)

0.391

10175 CNIH1 cornichon family AMPA receptor auxiliary protein 1

0.391

23555 TSPAN15

tetraspanin 15 0.391

8714 ABCC3 ATP-binding cassette, sub-family C (CFTR/MRP), member 3

0.390

8800 PEX11A

peroxisomal biogenesis factor 11 alpha 0.389

339745 SPOPL speckle-type POZ protein-like 0.389 3964 LGALS

8 lectin, galactoside-binding, soluble, 8 0.388

2796 GNRH1 gonadotropin-releasing hormone 1 (luteinizing-releasing hormone)

-0.388

400451 FAM174B

family with sequence similarity 174, member B

0.388

347735 SERINC2

serine incorporator 2 0.388

253782 CERS6 ceramide synthase 6 0.386 51702 PADI3 peptidyl arginine deiminase, type III 0.386 1362 CPD carboxypeptidase D 0.386 63027 SLC22

A23 solute carrier family 22, member 23 0.386

84311 MRPL45

mitochondrial ribosomal protein L45 0.386

401546 C9orf152

chromosome 9 open reading frame 152 0.385

57720 GPR107 G protein-coupled receptor 107 0.385 90060 CCDC1

20 coiled-coil domain containing 120 0.384

90861 HN1L hematological and neurological expressed 1-like

0.384

83606 GUCD1 guanylyl cyclase domain containing 1 0.384 30837 SOCS7 suppressor of cytokine signaling 7 0.383 100132074

FOXO6 forkhead box O6 0.383

54883 CWC25 CWC25 spliceosome-associated protein homolog

0.383

150696 PROM2 prominin 2 0.383

Page 60: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

53

NCBI ID

HUGO Gene Symbol

Gene Name r

55876 GSDMB

gasdermin B 0.383

6309 SC5D sterol-C5-desaturase 0.382 51109 RDH11 retinol dehydrogenase 11 (all-trans/9-

cis/11-cis) 0.382

51809 GALNT7

polypeptide N-acetylgalactosaminyltransferase 7

0.382

9618 TRAF4 TNF receptor-associated factor 4 0.381 4598 MVK mevalonate kinase 0.381 23505 TMEM1

31 transmembrane protein 131 0.381

83733 SLC25A18

solute carrier family 25 (glutamate carrier), member 18

0.381

116238 TLCD1 TLC domain containing 1 0.381 79739 TTLL7 tubulin tyrosine ligase-like family member

7 0.381

54793 KCTD9 potassium channel tetramerization domain containing 9

-0.381

134147 CMBL carboxymethylenebutenolidase homolog (Pseudomonas)

0.381

54432 YIPF1 Yip1 domain family member 1 0.380 7905 REEP5 receptor accessory protein 5 0.380 623 BDKRB

1 bradykinin receptor B1 0.379

9960 USP3 ubiquitin specific peptidase 3 0.379 80352 RNF39 ring finger protein 39 0.379 160418 TMTC3 transmembrane and tetratricopeptide repeat

containing 3 0.378

7587 ZNF37A

zinc finger protein 37A -0.378

1048 CEACAM5

carcinoembryonic antigen-related cell adhesion molecule 5

0.377

56649 TMPRSS4

transmembrane protease, serine 4 0.377

10321 CRISP3 cysteine-rich secretory protein 3 0.377 79618 HMBO

X1 homeobox containing 1 -0.377

254778 C8orf46 chromosome 8 open reading frame 46 -0.376 84074 QRICH

2 glutamine rich 2 -0.376

84059 ADGRV1

adhesion G protein-coupled receptor V1 0.376

91227 GGTLC2

gamma-glutamyltransferase light chain 2 0.375

Page 61: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

54

NCBI ID

HUGO Gene Symbol

Gene Name r

55664 CDC37L1

cell division cycle 37-like 1 -0.375

79813 EHMT1 euchromatic histone-lysine N-methyltransferase 1

0.375

26280 IL1RAPL2

interleukin 1 receptor accessory protein-like 2

0.374

100506144

ZMYM6NB

ZMYM6 neighbor -0.373

57551 TAOK1 TAO kinase 1 0.373 7351 UCP2 uncoupling protein 2 (mitochondrial, proton

carrier) 0.373

404093 CUEDC1

CUE domain containing 1 0.373

2296 FOXC1 forkhead box C1 -0.373 84640 USP38 ubiquitin specific peptidase 38 0.372 23229 ARHGE

F9 Cdc42 guanine nucleotide exchange factor 9

-0.372

5753 PTK6 protein tyrosine kinase 6 0.372 56929 FEM1C fem-1 homolog c (C. elegans) 0.372 389072 PLEKH

M3 pleckstrin homology domain containing, family M, member 3

-0.372

54884 RETSAT

retinol saturase (all-trans-retinol 13,14-reductase)

0.372

1213 CLTC clathrin, heavy chain (Hc) 0.372 286262 TPRN taperin 0.372 94160 ABCC1

2 ATP-binding cassette, sub-family C (CFTR/MRP), member 12

0.371

28983 TMPRSS11E

transmembrane protease, serine 11E 0.371

347475 CCDC160

coiled-coil domain containing 160 0.371

642273 FAM110C

family with sequence similarity 110, member C

0.371

23559 WBP1 WW domain binding protein 1 0.370 57649 PHF12 PHD finger protein 12 0.369 638 BIK BCL2-interacting killer (apoptosis-

inducing) 0.369

51263 MRPL30

mitochondrial ribosomal protein L30 0.368

29109 FHOD1 formin homology 2 domain containing 1 0.368 123036 TC2N tandem C2 domains, nuclear 0.368 55681 SCYL2 SCY1-like, kinase-like 2 0.368 23475 QPRT quinolinate phosphoribosyltransferase 0.368

Page 62: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

55

NCBI ID

HUGO Gene Symbol

Gene Name r

55568 GALNT10

polypeptide N-acetylgalactosaminyltransferase 10

0.368

80045 GPR157 G protein-coupled receptor 157 0.368 23677 SH3BP4 SH3-domain binding protein 4 0.367 3597 IL13RA

1 interleukin 13 receptor, alpha 1 0.367

79170 PRR15L proline rich 15-like 0.366 9817 KEAP1 kelch-like ECH-associated protein 1 -0.366 201266 SLC39

A11 solute carrier family 39, member 11 0.366

10200 MPHOSPH6

M-phase phosphoprotein 6 0.366

60592 SCOC short coiled-coil protein 0.365 91 ACVR1

B activin A receptor type IB 0.365

81576 CCDC130

coiled-coil domain containing 130 -0.365

124872 B4GALNT2

beta-1,4-N-acetyl-galactosaminyl transferase 2

0.365

6799 SULT1A2

sulfotransferase family 1A member 2 0.364

26996 GPR160 G protein-coupled receptor 160 0.364 84985 FAM83

A family with sequence similarity 83, member A

0.364

83451 ABHD11

abhydrolase domain containing 11 0.364

8464 SUPT3H

SPT3 homolog, SAGA and STAGA complex component

-0.363

51260 PBDC1 polysaccharide biosynthesis domain containing 1

0.363

122945 NOXRED1

NADP-dependent oxidoreductase domain containing 1

0.363

2140 EYA3 EYA transcriptional coactivator and phosphatase 3

-0.363

379 ARL4D ADP-ribosylation factor like GTPase 4D 0.363 9862 MED24 mediator complex subunit 24 0.363 9517 SPTLC2 serine palmitoyltransferase, long chain base

subunit 2 0.362

55040 EPN3 epsin 3 0.362 10509 SEMA4

B sema domain, immunoglobulin domain (Ig), transmembrane domain (TM) and short cytoplasmic domain, (semaphorin) 4B

0.361

51128 SAR1B secretion associated, Ras related GTPase 1B

0.361

Page 63: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

56

NCBI ID

HUGO Gene Symbol

Gene Name r

203413 CT83 cancer/testis antigen 83 -0.361 340719 NANOS

1 nanos homolog 1 (Drosophila) 0.361

55450 CAMK2N1

calcium/calmodulin-dependent protein kinase II inhibitor 1

0.361

255189 PLA2G4F

phospholipase A2, group IVF 0.361

222171 PRR15 proline rich 15 0.360 10299 MARC

H6 membrane-associated ring finger (C3HC4) 6, E3 ubiquitin protein ligase

0.360

253190 SERHL2

serine hydrolase-like 2 0.360

4850 CNOT4 CCR4-NOT transcription complex, subunit 4

-0.359

619189 SERINC4

serine incorporator 4 0.359

954 ENTPD2

ectonucleoside triphosphate diphosphohydrolase 2

0.359

57035 RSRP1 arginine/serine-rich protein 1 -0.359 148327 CREB3

L4 cAMP responsive element binding protein 3-like 4

0.359

9061 PAPSS1 3'-phosphoadenosine 5'-phosphosulfate synthase 1

-0.359

23583 SMUG1 single-strand-selective monofunctional uracil-DNA glycosylase 1

0.359

9630 GNA14 guanine nucleotide binding protein (G protein), alpha 14

0.358

168544 ZNF467 zinc finger protein 467 0.358 79675 FASTK

D1 FAST kinase domains 1 -0.357

367 AR androgen receptor 0.357 284340 CXCL1

7 chemokine (C-X-C motif) ligand 17 0.357

5429 POLH polymerase (DNA directed), eta -0.357 84216 TMEM1

17 transmembrane protein 117 -0.357

31 ACACA acetyl-CoA carboxylase alpha 0.356 57761 TRIB3 tribbles pseudokinase 3 0.356 201456 FBXO1

5 F-box protein 15 0.356

283377 SPRYD4

SPRY domain containing 4 0.356

5465 PPARA peroxisome proliferator-activated receptor alpha

-0.356

Page 64: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

57

NCBI ID

HUGO Gene Symbol

Gene Name r

7108 TM7SF2

transmembrane 7 superfamily member 2 0.356

10867 TSPAN9

tetraspanin 9 0.355

22928 SEPHS2 selenophosphate synthetase 2 0.355 5810 RAD1 RAD1 checkpoint DNA exonuclease 0.354 162979 ZNF296 zinc finger protein 296 0.354 51504 TRMT1

12 tRNA methyltransferase 11-2 homolog (S. cerevisiae)

0.354

22866 CNKSR2

connector enhancer of kinase suppressor of Ras 2

-0.354

598 BCL2L1

BCL2-like 1 0.353

8895 CPNE3 copine III 0.353 54875 CNTLN centlein, centrosomal protein -0.353 401474 SAMD1

2 sterile alpha motif domain containing 12 0.353

5829 PXN paxillin 0.353 671 BPI bactericidal/permeability-increasing protein -0.353 2244 FGB fibrinogen beta chain 0.353 79720 VPS37B vacuolar protein sorting 37 homolog B (S.

cerevisiae) 0.353

Page 65: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

58

Appendix 4. Gene lists (n = 116) which correlate with microcalcification (no,

low-to-intermediate, highly suspicious) and correlation coefficient (r) > 0.3

and false discovery rate < 0.05.

NCBI ID

HUGO Gene Symbol

Gene Name r

3038 HAS3 hyaluronan synthase 3 0.410 64175 P3H1 prolyl 3-hydroxylase 1 -0.398 338557 FFAR4 free fatty acid receptor 4 -0.383 94103 ORMD

L3 ORMDL sphingolipid biosynthesis regulator 3

0.380

29108 PYCARD

PYD and CARD domain containing -0.378

84624 FNDC1 fibronectin type III domain containing 1 -0.377 54894 RNF43 ring finger protein 43 0.361 5328 PLAU plasminogen activator, urokinase -0.358 1301 COL11

A1 collagen, type XI, alpha 1 -0.356

822 CAPG capping protein (actin filament), gelsolin-like

-0.355

100506144

ZMYM6NB

ZMYM6 neighbor -0.352

80896 NPL N-acetylneuraminate pyruvate lyase (dihydrodipicolinate synthase)

-0.351

5996 RGS1 regulator of G-protein signaling 1 -0.351 219 ALDH1

B1 aldehyde dehydrogenase 1 family, member B1

-0.351

2064 ERBB2 erb-b2 receptor tyrosine kinase 2 0.348 8654 PDE5A phosphodiesterase 5A, cGMP-specific 0.347 7805 LAPTM

5 lysosomal protein transmembrane 5 -0.347

713 C1QB complement component 1, q subcomponent, B chain

-0.347

2207 FCER1G

Fc fragment of IgE, high affinity I, receptor for; gamma polypeptide

-0.347

6507 SLC1A3

solute carrier family 1 (glial high affinity glutamate transporter), member 3

-0.345

284161 GDPD1 glycerophosphodiester phosphodiesterase domain containing 1

0.345

55080 TAPBPL

TAP binding protein-like -0.344

8828 NRP2 neuropilin 2 -0.342

Page 66: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

59

NCBI ID

HUGO Gene Symbol

Gene Name r

7045 TGFBI transforming growth factor, beta-induced, 68kDa

-0.342

113451 AZIN2 antizyme inhibitor 2 -0.342 2882 GPX7 glutathione peroxidase 7 -0.341 1501 CTNND

2 catenin (cadherin-associated protein), delta 2

0.341

2706 GJB2 gap junction protein beta 2 -0.341 2517 FUCA1 fucosidase, alpha-L- 1, tissue -0.341 586 BCAT1 branched chain amino-acid transaminase 1,

cytosolic -0.341

79899 PRR5L proline rich 5 like -0.340 639 PRDM1 PR domain containing 1, with ZNF domain -0.338 91614 DEPDC

7 DEP domain containing 7 -0.337

5329 PLAUR plasminogen activator, urokinase receptor -0.336 4804 NGFR nerve growth factor receptor 0.335 27244 SESN1 sestrin 1 0.334 4973 OLR1 oxidized low density lipoprotein (lectin-

like) receptor 1 -0.334

10437 IFI30 interferon, gamma-inducible protein 30 -0.334 79783 SUGCT succinyl-CoA:glutarate-CoA transferase -0.334 860 RUNX2 runt-related transcription factor 2 -0.333 339745 SPOPL speckle-type POZ protein-like 0.332 54407 SLC38

A2 solute carrier family 38, member 2 0.332

10082 GPC6 glypican 6 -0.332 2590 GALNT

2 polypeptide N-acetylgalactosaminyltransferase 2

-0.330

9618 TRAF4 TNF receptor-associated factor 4 0.329 7226 TRPM2 transient receptor potential cation channel,

subfamily M, member 2 -0.328

199 AIF1 allograft inflammatory factor 1 -0.328 220441 RNF152 ring finger protein 152 0.328 1508 CTSB cathepsin B -0.327 2745 GLRX glutaredoxin (thioltransferase) -0.326 1182 CLCN3 chloride channel, voltage-sensitive 3 0.326 2335 FN1 fibronectin 1 -0.326 50515 CHST11 carbohydrate (chondroitin 4)

sulfotransferase 11 -0.325

59351 PBOV1 prostate and breast cancer overexpressed 1 0.325 2886 GRB7 growth factor receptor-bound protein 7 0.325 154661 RUNDC

3B RUN domain containing 3B 0.324

57194 ATP10A ATPase, class V, type 10A -0.324

Page 67: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

60

NCBI ID

HUGO Gene Symbol

Gene Name r

11031 RAB31 RAB31, member RAS oncogene family -0.324 7515 XRCC1 X-ray repair complementing defective

repair in Chinese hamster cells 1 -0.324

51300 TIMMDC1

translocase of inner mitochondrial membrane domain containing 1

-0.323

5783 PTPN13 protein tyrosine phosphatase, non-receptor type 13 (APO-1/CD95 (Fas)-associated phosphatase)

0.322

83699 SH3BGRL2

SH3 domain binding glutamate-rich protein like 2

0.321

285180 RUFY4 RUN and FYVE domain containing 4 -0.321 3107 HLA-C major histocompatibility complex, class I,

C -0.321

729220 FLJ45513

uncharacterized LOC729220 0.320

23598 PATZ1 POZ (BTB) and AT hook containing zinc finger 1

0.320

115908 CTHRC1

collagen triple helix repeat containing 1 -0.320

3983 ABLIM1

actin binding LIM protein 1 0.320

84299 MIEN1 migration and invasion enhancer 1 0.319 5480 PPIC peptidylprolyl isomerase C (cyclophilin C) -0.318 968 CD68 CD68 molecule -0.318 25903 OLFML

2B olfactomedin like 2B -0.317

3652 IPP intracisternal A particle-promoted polypeptide

0.317

399664 MEX3D mex-3 RNA binding family member D 0.316 93210 PGAP3 post-GPI attachment to proteins 3 0.316 131578 LRRC1

5 leucine rich repeat containing 15 -0.314

6426 SRSF1 serine/arginine-rich splicing factor 1 0.313 8832 CD84 CD84 molecule -0.313 2825 GPR1 G protein-coupled receptor 1 -0.313 8590 OR6A2 olfactory receptor, family 6, subfamily A,

member 2 -0.313

10797 MTHFD2

methylenetetrahydrofolate dehydrogenase (NADP+ dependent) 2, methenyltetrahydrofolate cyclohydrolase

0.312

23384 SPECC1L

sperm antigen with calponin homology and coiled-coil domains 1-like

0.312

10855 HPSE heparanase -0.312

Page 68: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

61

NCBI ID

HUGO Gene Symbol

Gene Name r

2212 FCGR2A

Fc fragment of IgG, low affinity IIa, receptor (CD32)

-0.311

8452 CUL3 cullin 3 0.311 376267 RAB15 RAB15, member RAS oncogene family 0.311 54532 USP53 ubiquitin specific peptidase 53 0.311 114784 CSMD2 CUB and Sushi multiple domains 2 -0.311 79762 C1orf11

5 chromosome 1 open reading frame 115 0.311

157638 FAM84B

family with sequence similarity 84, member B

0.310

100506374

LOC100506374

uncharacterized LOC100506374 0.310

84640 USP38 ubiquitin specific peptidase 38 0.310 4323 MMP14 matrix metallopeptidase 14 (membrane-

inserted) -0.309

51705 EMCN endomucin 0.309 6404 SELPL

G selectin P ligand -0.309

6567 SLC16A2

solute carrier family 16, member 2 (thyroid hormone transporter)

0.309

1215 CMA1 chymase 1, mast cell 0.308 2532 ACKR1 atypical chemokine receptor 1 (Duffy blood

group) 0.308

55337 C19orf66

chromosome 19 open reading frame 66 -0.308

124801 LSM12 LSM12 homolog -0.308 26045 LRRTM

2 leucine rich repeat transmembrane neuronal 2

0.307

196051 PLPP4 phospholipid phosphatase 4 -0.307 441272 SPDYE

3 speedy/RINGO cell cycle regulator family member E3

0.307

90865 IL33 interleukin 33 0.307 440021 KRTAP

5-2 keratin associated protein 5-2 0.306

57609 DIP2B disco-interacting protein 2 homolog B 0.306 23411 SIRT1 sirtuin 1 0.306 692094 MSMP microseminoprotein, prostate associated 0.305 6578 SLCO2

A1 solute carrier organic anion transporter family, member 2A1

0.305

55083 KIF26B kinesin family member 26B -0.304 56889 TM9SF

3 transmembrane 9 superfamily member 3 0.304

8897 MTMR3

myotubularin related protein 3 0.304

Page 69: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

62

NCBI ID

HUGO Gene Symbol

Gene Name r

3021 H3F3B H3 histone, family 3B (H3.3B) 0.304 81832 NETO1 neuropilin (NRP) and tolloid (TLL)-like 1 -0.304 23505 TMEM1

31 transmembrane protein 131 0.304

1573 CYP2J2 cytochrome P450, family 2, subfamily J, polypeptide 2

0.300

Page 70: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

63

Appendix 5. Functional annotation clusters using DEGs (n = 1047) from highly suspicious/no microcalcifications comparison set in HER2-

negative subgroup which contain GO terms in the biological process category with enrichment score > 3.5

Gene Ontology ID Term Fold Enrichment

P Value Bonferroni

Annotation Cluster 1 Enrichment Score: 13.606 GO:0006952 defense response 3.317 1.73E-17 4.58E-14 GO:0006954 inflammatory response 3.863 7.48E-13 1.99E-09 GO:0009611 response to wounding 3.080 1.18E-12 3.13E-09 Annotation Cluster 2 Enrichment Score: 7.378 GO:0048002 antigen processing and presentation of peptide antigen 15.694 6.86E-13 1.82E-09 GO:0002504 antigen processing and presentation of peptide or

polysaccharide antigen via MHC class II 11.414 3.03E-09 8.06E-06

GO:0002478 antigen processing and presentation of exogenous peptide antigen

19.974 4.02E-07 1.07E-03

GO:0019886 antigen processing and presentation of exogenous peptide antigen via MHC class II

23.541 1.64E-06 4.35E-03

GO:0002495 antigen processing and presentation of peptide antigen via MHC class II

23.541 1.64E-06 4.35E-03

GO:0019884 antigen processing and presentation of exogenous antigen 15.694 2.41E-06 6.37E-03 Annotation Cluster 3 Enrichment Score: 7.159 GO:0042330 taxis 4.512 6.98E-09 1.85E-05 GO:0006935 chemotaxis 4.512 6.98E-09 1.85E-05 GO:0007626 locomotory behavior 3.207 1.88E-07 4.99E-04 GO:0007610 behavior 2.409 2.52E-06 0.007

Page 71: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

64

Gene Ontology ID Term Fold Enrichment

P Value Bonferroni

Annotation Cluster 4 Enrichment Score: 6.559 GO:0002684 positive regulation of immune system process 4.484 8.52E-13 2.26E-09 GO:0050778 positive regulation of immune response 4.979 1.05E-09 2.80E-06 GO:0002443 leukocyte mediated immunity 5.840 7.39E-08 1.96E-04 GO:0002252 immune effector process 4.450 2.38E-07 6.31E-04 GO:0048584 positive regulation of response to stimulus 3.325 5.13E-07 0.001 GO:0002250 adaptive immune response 5.707 8.14E-07 0.002 GO:0002460 adaptive immune response based on somatic

recombination of immune receptors built from immunoglobulin superfamily domains

5.707 8.14E-07 0.002

GO:0016064 immunoglobulin mediated immune response 6.394 6.48E-06 0.017 GO:0019724 B cell mediated immunity 6.165 9.11E-06 0.024 GO:0002449 lymphocyte mediated immunity 5.381 1.18E-05 0.031 GO:0002253 activation of immune response 4.007 1.89E-04 0.395 Annotation Cluster 5 Enrichment Score: 5.853 GO:0001817 regulation of cytokine production 4.335 3.00E-09 7.97E-06 GO:0001819 positive regulation of cytokine production 4.534 2.65E-05 0.068 GO:0051240 positive regulation of multicellular organismal process 2.830 3.48E-05 0.088 Annotation Cluster 6 Enrichment Score: 4.247 GO:0006928 cell motion 2.313 8.58E-06 0.023 GO:0016477 cell migration 2.616 7.24E-05 0.175 GO:0051674 localization of cell 2.454 1.29E-04 0.290 GO:0048870~ cell motility 2.454 1.29E-04 0.290 Annotation Cluster 7 Enrichment Score: 3.920

Page 72: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

65

Gene Ontology ID Term Fold Enrichment

P Value Bonferroni

GO:0002237 response to molecule of bacterial origin 4.745 1.66E-05 0.043 GO:0009617 response to bacterium 2.927 1.45E-04 0.319 GO:0032496 response to lipopolysaccharide 4.076 7.25E-04 0.854 Annotation Cluster 8 Enrichment Score: 3.702 GO:0045321 leukocyte activation 3.243 8.10E-07 0.002 GO:0001775 cell activation 2.843 5.12E-06 0.013 GO:0042110 T cell activation 3.487 1.85E-04 0.388 GO:0046649 lymphocyte activation 2.524 0.002 0.990 GO:0030098 lymphocyte differentiation 1.828 0.230 1

Page 73: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

66

(국 )

: 에 미 회 는 암 조 견하고

진단하는데 요하다. 하지만 암 회 후 에 한

생 학 에 해 는 진 가 없다. 본 연구에 는

마이크 어 이 분 법 이용하여, 에 미 회

자 계를 알아보는 것 목 하 다.

법: 168 명 암 자(나이 앙값, 50 ; 범 , 21-

79 )에 Affymetrix GeneChip® Human Gene 2.0 ST

어 이(53,617 프 )를 이용한 마이크 어 이 데이 를

획득했다. 명 상 학과 사가 상 미 회

소견 분 하여, 미 회 가 없는 그룹(n=99), 등도

악 가능 그룹(n=37), 높 악 가능 그룹(n=32)

분 하 다. 각 그룹간 차별 자를 별하 하여

일원 치 분산분 시행하 고, 키법 이용한 사후분

시행하여 P value < 0.05 를 용하 다. 차별 자들

생 학 인 미를 찾아내 하여 자 토 지 군집

분 시행하 다.

결과: 3 군 에 2551 개 자가 에 차이가

있었다. 높 악 가능 /미 회 없는 그룹 에 는

1838 개 자(과 955 개, 883 개), 높

악 가능 / 등도 악 가능 그룹 에 는 484 개

Page 74: Disclaimer - Seoul National Universitys-space.snu.ac.kr/bitstream/10371/121828/1/000000135935.pdfframework based on concept lattice analysis to make association of gene expression

67

자(과 342 개, 142 개), 등도

악 가능 /미 회 없는 그룹 에 는 457 개

자(과 126 개, 331 개)가 견 었다. 자

토 지 군집 분 통하여 높 악 가능 회 그룹에

미 회 가 없는 그룹에 해 면역, 어 염증 이

감소 어 있었 며 특히 인간 피 장인자 용체 2- 인

그룹에 면역시스 이 미 회 강한 이

있었다.

결 : 암에 자 미 회

상태에 라 상이하다. 미 회 를 포함하는 암

면역시스 감소 이 있다.

-------------------------------------

주요어: 암, 미 회 , 마이크 어 이,

인간 피 장인자 용체 2, 면역시스

학번: 2014-30911