Constructing Gene Expression Modules

We defined modules as sets of co-expressed genes that were considered as a functional unit. Using multiple approaches, we built a collection of 323 gene expression modules, including 115 gene lists obtained from 53 publications: (1) 221 modules were built using the median expression of all genes within the module that homogeneously expresses these genes (i.e. all genes in the module were high or low together within a given sample). The sources of the selected homogenous gene lists were the following: 50 were identified by bicluster analyses[54 (link)] using the microarray dataset of 359 human breast tumors and 8 normal breast samples (i.e. the aforementioned 2/3 training set); 52 modules were identified from an unsupervised hierarchical clustering analysis of the same human breast tumor database; 50 were identified by bicluster analyses using microarray data of 122 mouse mammary tumors[13 (link)]; 56 were identified from unsupervised hierarchical clustering analysis of the same mouse mammary database; 13 were identified from previously published gene lists [13 (link),14 (link),17 (link),18 (link),21 (link),26 (link),35 (link),55 (link)]. (2) 77 modules were represented as the first Principal Component of previously published gene lists [3 (link),8 (link)-10 (link),12 (link),13 (link),15 (link),19 (link),20 (link),22 (link),23 (link),25 (link),27 (link)-32 (link),34 (link),36 (link),37 (link),40 (link),56 (link)-66 (link)] that showed heterogeneous expression patterns (i.e. the gene list contained genes with high and low expression within a given sample). (3) 22 modules were correlations to previously published training dataset centroids [4 (link),9 (link),11 (link),16 (link),24 (link),33 (link),38 (link),39 (link),50 (link),67 (link),68 (link)]. (4) 3 modules were built from previously published gene expression prognostic models [5 (link),46 (link),47 (link)]. We acknowledge that our implementation of some of the previously published signatures may be suboptimal, however, we attempted within reason, to apply each signature as published. All modules, with gene lists and references, can be found in Additional File 1.

Free full text: Click here

Fan C., Prat A., Parker J.S., Liu Y., Carey L.A., Troester M.A, & Perou C.M. (2011). Building prognostic models for breast cancer patients using clinical variables and hundreds of gene expression signatures. BMC Medical Genomics, 4, 3.

Publication 2011

A gene Gene Gene component Gene expression Genes module Heterogeneous Homogenous Human Mammary Mammary tumors Microarray Mouse Tumor of the breast

Corresponding Organization :

Other organizations : University of North Carolina at Chapel Hill

Top 5 similar protocols

Protocol cited in 31 other protocols

Variable analysis

independent variables

Homogeneity of gene expression within modules
Previously published gene lists and centroids
Previously published gene expression prognostic models

dependent variables

Collection of 323 gene expression modules

control variables

None explicitly mentioned

controls

Positive controls: None mentioned
Negative controls: None mentioned

Annotations

Based on most similar protocols

Etiam vel ipsum. Morbi facilisis vestibulum nisl. Praesent cursus laoreet felis. Integer adipiscing pretium orci. Nulla facilisi. Quisque posuere bibendum purus. Nulla quam mauris, cursus eget, convallis ac, molestie non, enim. Aliquam congue. Quisque sagittis nonummy sapien. Proin molestie sem vitae urna. Maecenas lorem.

As authors may omit details in methods from publication, our AI will look for missing critical information across the 5 most similar protocols.

About PubCompare

Our mission is to provide scientists with the largest repository of trustworthy protocols and intelligent analytical tools, thereby offering them extensive information to design robust protocols aimed at minimizing the risk of failures.

We believe that the most crucial aspect is to grant scientists access to a wide range of reliable sources and new useful tools that surpass human capabilities.

However, we trust in allowing scientists to determine how to construct their own protocols based on this information, as they are the experts in their field.

Ready to get started?

Revolutionizing how scientists
search and build protocols!