Skip to content

shbrief/model_building

Repository files navigation

Model building process

Scripts here are used to build RAVmodel_536, annotated with MSigDB C2.all gene sets.

Script Processes Main Output
01_import.R Import _quant.sf files using tximport {studyName}.rds
02_Filtering.R Log transformation {studyName}_count.csv
03_Top_Genes.R Order genes in each study based on their variance topGenesInTrainingData
04_Common_Genes.R Select top-varing common genes topGenes
05_PCA.R Calculate mean and sd for all samples/ rowNormalization/ PCA _PCs_rowNorm.rds
06_Clustering.R Clustering/ Build RAVindex/ Calculate avg.silhouette.width _PCclusters.rds
07_Final_Model.R MeSH/GSEA annotation and build RAVmodel _RAVmodel_{annotGeneSets}.rds

A list of RAVmodels

Here is the summary table of studies/samples/genes for different models.

# of studies # of samples % top varying # of genes
536 44,890 90% 13,934
735 51,605 90% 11,685
1,399 75,433 90% 7,951

About

RAVmodel Building Workflows

Topics

Resources

Stars

Watchers

Forks

Packages

 
 
 

Contributors