Using expression profiling to identify human microRNA targets

Using expression profiling to identify human microRNA targets

Jim C. Huang^1,6, Tomas Babak^2,6, Timothy W. Corson^2,3, Gordon Chua⁴, Sofia Khan^2,3, Brenda L. Gallie^2,3, Timothy R. Hughes^2,4,
Benjamin J. Blencowe^2,4, Brendan J. Frey^1,4,5,7 and Quaid D. Morris^2,4,5,7

¹ Department of Electrical and Computer Engineering, University of Toronto, 10 King's College Rd., Toronto, ON, M5S 3G4, Canada
² Department of Molecular and Medical Genetics, University of Toronto, 1 King's College Rd., Toronto, ON, M5S 1A8, Canada
³ Division of Applied Oncology, Ontario Cancer Institute/Princess Margaret Hospital, University Health Network, Toronto, ON, M5G 2M9, Canada
⁴ Banting and Best Department of Medical Research, University of Toronto, Toronto, ON, 160 College St., M5G 1L6, Canada
⁵ Department of Computer Science, University of Toronto, 10 King's College Rd., M5S 3G4, Canada
⁶ These authors contributed to this work equally
⁷ To whom correspondence should be addressed: frey@psi.toronto.edu, quaid.morris@utoronto.ca

Frequently asked questions

Q: Where can I find the expression data sets used in the paper?
A: Microarray data sets are available at NCBI GEO (GEO Accession: GSE7072 (microRNA) and GSE7185 (mRNA)). The GSE7072 data set is the microRNA expression data shown in Figure 2A of the paper. The GSE7185 data set is the mRNA expression data used in tandem with mRNA sequence to find enriched motifs and generate Figure 3A of the paper.
The mRNA and microRNA expression data sets used to find functional microRNA-target interactions can be downloaded in MATLAB format (requires MATLAB v7.0 or above) at www.psi.toronto.edu/genmir/code/. The files are provided sparately or together in one zip file. The mRNA and miRNA expression data sets used to find functional miRNA-target interactions is miRdata_Golub_TSS_human.mat. The output of GenMiR++ is stored in GenMiRout_Golub_TSS_human.mat.

Q: What do the fields in the MATLAB file for the mRNA and miRNA expression data correspond to?
A: The MATLAB files contain these different fields:

M: Number of microRNAs (114)
N: Number of mRNAs (16063)
T: Number of tissue/cell types (88)
tissues: Tissue labels
X: NxT matrix of mRNA expression profiles (one per row)
glbl_HGNC,glbl_EMBL: mRNA identifiers in HGNC and EMBL format
Z: MxT matrix of microRNA expression profiles (one per row)
mlbl: microRNA identifiers
C: MxN binary matrix of indicators of whether a given microRNA putatively targets a given mRNA
Beta: MxN matrix of posterior probabilities of whether a given microRNA putatively targets a given mRNA under the GenMiR++ model

Other fields are explained in this Supplementary Description.

Q: I'd like to re-run the analyses presented in the paper. How do I go about it?
A: In order to reproduce the results and plots shown in the paper, please follow these steps:

Ensure you have MATLAB v7.0 or above installed on your system (may require administrator rights);
Download all MATLAB functions and scripts of www.psi.toronto.edu/genmir/code/ to your local system. These are provided both as separate .m files and together as one zip file (GenMiR_code_data.zip);
To reproduce the set of functional microRNA-target interactions, re-run the GenMiR++ algorithm on the ./data/miRdata_Golub_TSS_human.mat data set. The network of microRNA-target interactions corresponding to selecting the upper 25% of interactions is shown in Supplementary Figures 3A, 3B of the paper.
To reproduce Supplementary Figures 3A, 3B , run ./Fig3/GenMiR_GO_analysis_human.m (this requires the above GenMiR++ output file);
To reproduce the quantities in Figure 3A, run ./Fig5A/display_let7b_transfection_results.m (this requires the above GenMiR++ output file);
Motifs for which down-regulated mRNAs were enriched can be obtained in ./Fig5A/let7b_downregulated_motifs_*_new.out. Run BioProspector on the 3'UTR sequences found in ./Fig5A/let7b_downregulated.txt with the flags -r 25 -W w, where w is the desired motif length.

Page maintained by Jim Huang (jim@psi.toronto.edu)