Systematic identification of transcription factors associated with patient survival in cancers

Abstract

Background: Aberrant activation or expression of transcription factors has been implicated in the tumorigenesis of various types of cancer. In spite of the prevalent application of microarray experiments for profiling gene expression in cancer samples, they provide limited information regarding the activities of transcription factors. However, the association between transcription factors and cancers is largely dependent on the transcription regulatory activities rather than mRNA expression levels.

Results: In this paper, we propose a computational approach that integrates microarray expression data with the transcription factor binding site information to systematically identify transcription factors associated with patient survival given a specific cancer type. This approach was applied to two gene expression data sets for breast cancer and acute myeloid leukemia. We found that two transcription factor families, the steroid nuclear receptor family and the ATF/CREB family, are significantly correlated with the survival of patients with breast cancer; and that a transcription factor named T-cell acute lymphocytic leukemia 1 is significantly correlated with acute myeloid leukemia patient survival.

Conclusion: Our analysis identifies transcription factors associating with patient survival and provides insight into the regulatory mechanism underlying the breast cancer and leukemia. The transcription factors identified by our method are biologically meaningful and consistent with prior knowledge. As an insightful tool, this approach can also be applied to other microarray cancer data sets to help researchers better understand the intricate relationship between transcription factors and diseases.

Figure 1

Survival analysis of breast cancer subgroups defined based on activities of negative PWM predictors. The "AC>2" and "AC<-2" subgroups are defined based on the AC scores of V$PR_02 in (A), V$E2F_03 in (B), V$CREBP1_Q2 in (C), or V$AR_02 in (D). The survival curves are estimated using the Kaplan-Meier method and the difference between subgroups is examined by the log-rank test.

Figure 2

Survival analysis of breast cancer subgroups defined based on activities of positive PWM predictors. The "AC>2" and "AC<-2" subgroups are defined based on the AC scores of V$PAX9_B in (A) or V$LXR_DR4_Q3 in (B).

Figure 3

Survival analysis of subgroups defined based on PWM activities in ER-positive and ER-negative breast cancers. The "AC>2" and "AC<-2" subgroups are defined based on the AC scores of (A): V$PR_02 in ER-positive breast cancer, (B): V$PR_02 in ER-negative breast cancer, (C): V$CEBPDELTA_Q6 in ER-positive breast cancer, and (D): V$CEBPDELTA_Q6 in ER-negative breast cancer.

Figure 4

Survival analysis of subgroups defined based on PWM activities in ER-negative and ER-positive breast cancers. The "AC>2" and "AC<-2" subgroups are defined based on the AC scores of (A): V$ATF4_Q2 in ER-negative breast cancer, (B): V$ATF4_Q2 in ER-positive breast cancer, (C): V$ATF3_Q6 in ER-negative breast cancer, and (D): V$ATF3_Q6 in ER-positive breast cancer.

Figure 5

Survival analysis of AML subgroups defined based on PWM activities. The "AC>2" and "AC<-2" subgroups are defined based on the AC scores of V$TAL1BETAE47_01 in (A) or V$TAL1ALPHAE47_01 in (B). The "+" signs mark the events at which a sample is censored.