Merge pull request #2 from d3b-center/rerun-modules

Rerun modules

.gitignore

@@ -377,4 +377,7 @@ Icon
 .AppleDesktop
 Network Trash Folder
 Temporary Items
-.apdisk
+.apdisk
+
+# add data file to gitignore
+data/v15

analyses/data_preparation/.Renviron

@@ -1,2 +1 @@
 R_MAX_VSIZE=100Gb
-

analyses/data_preparation/.gitignore

@@ -1 +1,2 @@
-data/*
+# outputs
+results/*

analyses/data_preparation/01-multi-modal-clustering-prepare-data.R

@@ -7,6 +7,8 @@ suppressPackageStartupMessages({
   library(rtracklayer)
 })
 
+mem.maxVSize(vsize = 102400)
+
 # parse command line options
 option_list <- list(
   make_option(c("--histology_file"), type = "character", help = "Histology file (.tsv)"),

analyses/data_preparation/README.md

@@ -1,16 +1,14 @@
-
-### Author: Komal S. Rathi
+### Author: Komal S. Rathi, Adam Kraya
 
 ### Purpose
 
 The purpose of this module is:
-1. Subset input data matrices (RNA, CNV, SNV, Methylation and Splicing data) to short histology of interest.
-2. Format PSI matrix to a format accepted by PEGASAS.
-3. Filter/Transform data matrices which can be used as inputs for multi-modal clustering packages. 
+1. Subset input data matrices (RNA, CNV, SNV, Methylation and Splicing data) to short histology of interest i.e `HGAT`. 
+2. Filter/Transform data matrices which can be used as inputs for multi-modal clustering packages. 
 
 ### Data version
 
-- OPC v15 for histologies (molecular subtype etc), RNA, Methylation, Splicing
+- OPC v15 for histologies (with clinical variables like molecular subtype), RNA, Methylation, Splicing
 
 ### Run Analysis
 ```
@@ -29,29 +27,27 @@ Features were selected from OpenPedCan-analysis v15 datasets using the following
 
 1) **RNA**: 
 
-- The expected counts dataset was first filtered to Medulloblatoma samples. 
+- The expected counts dataset was first filtered to `HGAT` samples. 
 - Features were reduced to `Top 1000 most variable protein coding genes` followed by `Rank transformation`.
 
 2) **Methylation**:
 
-- Methylation beta-values matrix was first filtered to Medulloblatoma samples. 
+- Methylation beta-values matrix was first filtered to `HGAT` samples. 
 - Features were reduced to `Top 1000 most variable probes`.
 
 3) **Splicing**:
 
-- Splice matrix was first filtered to Medulloblatoma samples. 
+- Splice matrix was first filtered to `HGAT` samples. 
 - Features were reduced to `Top 1000 most variable splice variants`.
 
 #### Output
 
-The script resulted in `152 Medulloblastoma samples` that have all 3 data modalities available. The filtered/transformed data matrices are written out to individual .tsv files. Additionally a mapping between `Kids_First_Biospecimen_ID` identifiers from each modality and `sample_id` is written out to  `samples_map.tsv`.
+The script resulted in `228 HGAT samples` that have all 3 data modalities available. The filtered/transformed data matrices are written out to individual .tsv files. Additionally a mapping between `Kids_First_Biospecimen_ID` identifiers from each modality and `sample_id` is written out to  `samples_map.tsv`.
 
 ```
 results
-├── cnv_data.tsv # cnv data 
-├── methyl_data.tsv # methylation data 
-├── norm_counts.tsv # expression data 
-├── snv_data.tsv # snv data 
-├── splice_data.tsv # splice data 
+├── methyl_data.tsv # methylation data
+├── rna_data.tsv # expression data
+├── splice_data.tsv # splice data
 └── samples_map.tsv # biospecimens + cohort identifiers for samples used for each modality 
 ```

analyses/dge_pathway_analysis/.gitignore

@@ -0,0 +1,3 @@
+# outputs
+results/*
+plots/*

analyses/dge_pathway_analysis/README.md

@@ -17,11 +17,11 @@ bash run_analysis.sh
 #### Inputs
 
 ```
-../../data
-├── c2.cp.kegg_medicus.v2023.2.Hs.symbols.gmt # KEGG MEDICUS gmt file
-└── gencode.v39.primary_assembly.annotation.gtf.gz # gencode v39
+# gencode v39
+../../data/v15
+└── gencode.v39.primary_assembly.annotation.gtf.gz 
 
-# cohort specific files
+# gene expression file
 ../data_preparation/data
 └── gene-counts-rsem-expected_count-collapsed.rds
 
@@ -59,4 +59,3 @@ plots/intNMF/deseq
     ├── cluster_{n}_vs_rest_gsea_cnet.pdf
     └── cluster_{n}_vs_rest_gsea_dotplot.pdf
 ```
-