Dataset Information

Prototypes for 4x44K PAM50 Subtyping

ABSTRACT: PAM50 intrinsic subtype classification (Parker JS, Mullins M, Cheang MCU, et al: Supervised Risk Predictor of Breast Cancer Based on Intrinsic Subtypes. J Clin Oncol 27:1160-1167, 2009) was used for the Z1031 patient cohort. However, to account for potential technical biases such as differences in platforms and laboratory protocols between the Z1031 data set and the PAM50 training set2, we normalized the Z1301 data set before applying the PAM50 algorithm. This is a critical step that must be taken into consideration for appropriate subtype predictions as our group and others have previously shown3,4. To appropriately normalize the Z1301 microarray data, which is entirely composed of ER-positive tumors, we needed to estimate technical bias from a cohort with equivalent subtype representation as the PAM50 training set. A set including ER-negative, HER2-positive and true normal breast samples were profiled under the same platform, protocol and laboratory (i.e. WashU) as the Z1301 data set. A total of 40 prototypic samples of the various molecular subtypes were selected to match the distribution in the PAM50 training set. The mean expression of each PAM50 gene was calculated in this set of 40. These values obtained from the prototypical sample set were used for calibration purposes and were included as a new WashU_4X44K_Calibration column in the CalibrationParameter file, which has been already provided in the PAM50 algorithm2. This adjustment step assures that each sample is placed in the appropriate ‘prototypic space’ by subtracting the calibration value of each PAM50 gene from the expression value of the corresponding gene in each tested sample. Finally, single sample predictions were performed in every Z1301 sample using the WashU_4X44K_Calibration parameter. The complete data set for the 40 prototypic arrays is submitted under GSE26082. 40 experiments were used as prototypical cases for subtype calibration purposes.

ORGANISM(S): Homo sapiens

SUBMITTER: Ellis MJ

PROVIDER: S-ECPF-GEOD-26082 | biostudies-other | 2011 Jun

REPOSITORIES: biostudies-other

ACCESS DATA

Similar Datasets

Project description:BackgroundEarly detection of breast cancer in blood is both appealing clinically and challenging technically due to the disease's illusive nature and heterogeneity. Today, even though major breast cancer subtypes have been characterized, i.e., luminal A, luminal B, HER2+, and basal-like, little is known about the heterogeneity of breast cancer in blood, which could help to discover minimally invasive protein biomarkers with which clinical researchers can detect, classify, and monitor different breast cancer subtypes.ResultsIn this study, we performed an integrative pathway-assisted clustering analysis of breast cancer subtypes from plasma proteome samples collected from 80 patients diagnosed with breast cancer and 80 healthy women. First, four breast cancer subtypes and additionally unknown subtype (according to existing annotation) were determined based on pathology lab test results in primary tumors of enrolled patients. Next, we developed and applied four distance metrics, i.e., Protein Intensity, Q-Value, Pathway Profile, and Distance Score Function, to measure and characterize these cancer subtypes. Then, we developed a permutation test to evaluate the significant protein level changes in each biological pathway for each breast cancer subtype, using q-value. Lastly, we developed a pathway-protein matrix for each of the four distance methods to estimate the distance between breast cancer subtypes, for which further Pathway Association Network analysis were performed.ConclusionsWe found that 1) the luminal group (luminal A and luminal B) are clustered together, as well as the basal group (basal-like and HER2+) and 2) luminal A and luminal B are more close to each other than basal-like and HER2+ to each other. Our results were consistent with a recent independent breast cancer research from the Cancer Genome Atlas Network using genomic DNA copy number arrays, DNA methylation, exome sequencing, messenger RNA arrays, microRNA sequencing and reverse-phase protein arrays. Our results showed that changes of different breast cancer subtypes at the pathway level are more profound and less variable than those at the molecular level. Similar subtypes share distinct yet similar pathway activation networks, while dissimilar subtypes are different also at the level of pathway activation networks. The results also showed that distance or similarity of cancer subtypes based on pathway analysis might be able to provide further insight into the intrinsic relationship of breast cancer subtypes. We believe integrative pathway-assisted proteomics analysis described here can become a model for reliable clustering or classification of other cancer subtypes.

Project description:Breast cancer is commonly classified into intrinsic molecular subtypes. Standard gene centering is routinely done prior to molecular subtyping, but it can produce inaccurate classifications when the distribution of clinicopathological characteristics in the study cohort differs from that of the training cohort used to derive the classifier.We propose a subgroup-specific gene-centering method to perform molecular subtyping on a study cohort that has a skewed distribution of clinicopathological characteristics relative to the training cohort. On such a study cohort, we center each gene on a specified percentile, where the percentile is determined from a subgroup of the training cohort with clinicopathological characteristics similar to the study cohort. We demonstrate our method using the PAM50 classifier and its associated University of North Carolina (UNC) training cohort. We considered study cohorts with skewed clinicopathological characteristics, including subgroups composed of a single prototypic subtype of the UNC-PAM50 training cohort (n = 139), an external estrogen receptor (ER)-positive cohort (n = 48) and an external triple-negative cohort (n = 77).Subgroup-specific gene centering improved prediction performance with the accuracies between 77% and 100%, compared to accuracies between 17% and 33% from standard gene centering, when applied to the prototypic tumor subsets of the PAM50 training cohort. It reduced classification error rates on the ER-positive (11% versus 28%; P = 0.0389), the ER-negative (5% versus 41%; P < 0.0001) and the triple-negative (11% versus 56%; P = 0.1336) subgroups of the PAM50 training cohort. In addition, it produced higher accuracy for subtyping study cohorts composed of varying proportions of ER-positive versus ER-negative cases. Finally, it increased the percentage of assigned luminal subtypes on the external ER-positive cohort and basal-like subtype on the external triple-negative cohort.Gene centering is often necessary to accurately apply a molecular subtype classifier. Compared with standard gene centering, our proposed subgroup-specific gene centering produced more accurate molecular subtype assignments in a study cohort with skewed clinicopathological characteristics relative to the training cohort.

Dataset Information

Prototypes for 4x44K PAM50 Subtyping

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets