bhklab / metagx Goto Github PK

View Code? Open in Web Editor NEW

5.0 26.0 8.0 12.8 MB

Scripts to perform meta-analysis of cancer gene expression datasets

R 100.00%

gene-expression meta-analysis

metagx's Introduction

MetaGx

Scripts to perform meta-analysis of cancer gene expression datasets

Dependencies:

Install the R/Bioconductior dependencies:

pp <- c("Biobase", "BiocGenerics", "org.Hs.eg.db", "survival", "survcomp", "genefu", "mRMRe", "WriteXLS")

source("http://bioconductor.org/biocLite.R")

myrepos <- biocinstallRepos()

rr <- biocLite(pkgs=pp, dependencies=TRUE, type="source", destdir=".")

If using Windows:

WriteXLS needs Perl to be installed. To check if perl has been installed: library(WriteXLS) testPerl(perl="perl", verbose=TRUE)

TODO

Adapt the package to handle curatedOvarianData
Weighted survival does not work properly (number of patients per time points should be "unweighted")

metagx's People

Contributors

Stargazers

Watchers

Forkers

lavkan dgendoo wikum ankitshukla288 xulijunji zhangyupisa hossein-fallahi

metagx's Issues

Inconsistent annotations

The following annotations and corresponding column names must be include in each dataset in MetaGx and documented

PROBEID: probe name on the platform
ENTREZID: NCBI Entrez gene ID
ENSEMBLID: Ensembl gene id
SYMBOL: Official gene symbol

Check probeGeneMapping, subtypeClassification and all other functions

subscripts out of bound error

In robustness_validation.RNW, clustering.subtypes$tcga_nmf_4 is a list of length 15 as expected; but the elements of the lists are empty, while a list of at least 100 is expected

read.xls not imported NAMESPACE

As a result, getHellandSubtypes.R errors if library(gdata) is not done in global environment:

https://github.com/bhklab/MetaGx/blob/master/R/getHellandSubtypes.R#L2

In general, MetaGx has a heavy footprint and introduces risks of collisions with its many DEPENDS. Would be better to import functions, unless an entire package is really needed. E.g. see http://r-pkgs.had.co.nz/namespace.html - although Hadley's strict function-by-function importing may sometimes be overkill esp when using many functions from a base Bioconductor package, in many cases it is not much effort and avoids NAMESPACE collisions and attaching many unused functions and methods.

heavy use of DEPENDS

MetaGx has a heavy footprint and introduces risks of collisions with its many DEPENDS. Would be better to import functions, unless an entire package is really needed. E.g. see http://r-pkgs.had.co.nz/namespace.html - although Hadley's strict function-by-function importing may sometimes be overkill esp when using many functions from a base Bioconductor package, in many cases it is not much effort and avoids NAMESPACE collisions and attaching many unused functions and methods.

`select()` needs to be imported in `getHellandSubtypes()`

NAMESPACE collision with select from library(dplyr) is causing an error in OvcSubtypes classificationAcrossDatasets.Rnw (see reproducible example below, will attach file PMID17290060_eset.rds). This is related to Issue 3 - this error probably doesn't happen on an earlier version of R/Bioc/dplyr used during development, but breaks on more recent installations - such problems can be avoided by more use of imports rather than depends.

> suppressPackageStartupMessages(library(Biobase))
> suppressPackageStartupMessages(library(MetaGx))
> suppressPackageStartupMessages(library(gdata))
> suppressPackageStartupMessages(library(hgu133plus2.db))
> eset=readRDS("PMID17290060_eset.rds")
> tmp=getHellandSubtypes(eset)
'select()' returned 1:many mapping between keys and columns
'select()' returned 1:many mapping between keys and columns
'select()' returned 1:many mapping between keys and columns
'select()' returned 1:many mapping between keys and columns
> suppressPackageStartupMessages(library(dplyr))
> tmp=getHellandSubtypes(eset)
Error in UseMethod("select_") : 
  no applicable method for 'select_' applied to an object of class "c('ChipDb', 'AnnotationDb', 'envRefClass', '.environment', 'refClass', 'environment', 'refObject', 'AssayData')"
> sessionInfo()
R Under development (unstable) (2016-01-30 r70052)
Platform: x86_64-apple-darwin13.4.0 (64-bit)
Running under: OS X 10.10.5 (Yosemite)

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] stats4    parallel  stats     graphics  grDevices utils     datasets  methods  
[9] base     

other attached packages:
 [1] dplyr_0.4.3            hgu133plus2.db_3.2.2   gdata_2.17.0          
 [4] MetaGx_0.9.10          lsa_0.73.1             SnowballC_0.5.1       
 [7] WriteXLS_4.0.0         mRMRe_2.0.5            igraph_1.0.1          
[10] genefu_2.3.2           AIMS_1.3.0             e1071_1.6-7           
[13] iC10_1.1.3             iC10TrainingData_1.0.1 pamr_1.55             
[16] cluster_2.0.3          biomaRt_2.27.2         limma_3.27.11         
[19] mclust_5.1             survcomp_1.21.0        prodlim_1.5.7         
[22] survival_2.38-3        jetset_3.1.3           org.Hs.eg.db_3.2.3    
[25] RSQLite_1.0.0          DBI_0.3.1              AnnotationDbi_1.33.7  
[28] IRanges_2.5.24         S4Vectors_0.9.26       Biobase_2.31.3        
[31] BiocGenerics_0.17.3   

loaded via a namespace (and not attached):
 [1] Rcpp_0.12.3        bitops_1.0-6       class_7.3-14       tools_3.3.0       
 [5] bootstrap_2015.2   amap_0.8-14        gtools_3.5.0       SuppDists_1.1-9.2 
 [9] grid_3.3.0         R6_2.1.2           XML_3.98-1.3       lava_1.4.1        
[13] rmeta_2.16         magrittr_1.5       survivalROC_1.0.3  splines_3.3.0     
[17] assertthat_0.1     KernSmooth_2.23-15 RCurl_1.95-4.7    
>

bhklab / metagx Goto Github PK

metagx's Introduction

MetaGx

metagx's People

Contributors

Stargazers

Watchers

Forkers

metagx's Issues

Inconsistent annotations

subscripts out of bound error

read.xls not imported NAMESPACE

heavy use of DEPENDS

`select()` needs to be imported in `getHellandSubtypes()`

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent