jchen1981 / gmpr Goto Github PK

Hi there,

Just wondering whether GMPR is a contained R package that I should install (so far both devtools::install_github("https://github.com/jchen1981/GMPR") and devtools::install_local("~/Downloads/GMPR_0.1.3.tar.gz") returned errors. It would be great if you could update the README file with some installation instructions

Cheers

Hi,
I´trying to get the normalized counts after using GMPR.

...
dds <- DESeq(dds,parallel=TRUE)

# GMPR factors
otu.tab <- GMPR(counts(dds,norm=FALSE),intersect.no=10)
gmpr.size.factor <- otu.tab$gmpr
sizeFactors(dds) <-  gmpr.size.factor
dds <- estimateDispersions(dds)

# Get normalized
mdat <- counts(dds, normalized=TRUE)

Is the DESeq counts the best way to get normalized counts ?

Hello,
Could you give an example with some repdoucible data on how to use it ??

thanks.

Hi there,

I'm found your GMPR article very interesting. I'm trying to run the GMPR.R script for my otu.table, but I got this error:

gmpr.size.factor <- GMPR(otu_table)
Begin GMPR size factor calculation ...
Error in is.nan(pr) : default method not implemented for type 'list'
Called from: `[<-.data.frame`(`*tmp*`, is.nan(pr) | !is.finite(pr) | pr == 
    0, value = NA)

I check the type of both files: example data and my own and got "integer" and "list", respectively. But as far I can see the format of both files is identical. I'm new to R analysis, can you please help me?

Hello! Thank you for developing this easy to use normalization technique!

The size factors calculated between the provided C library and R script appear to be slightly different. Example code and results are below.

library(vegan)
library(GMPR) # C library
source('GMPR.R') # R script

data(throat.otu.tab)
data(throat.meta)

# Calculate GMPR size factor
# Row - features, column - samples

otu.tab <- t(throat.otu.tab)
gmpr.size.factor <- GMPR(otu.tab)
gmpr.size.factor.1 <- GMPR::GMPR(t(otu.tab))

> head(gmpr.size.factor)
 ESC_1.1_OPL  ESC_1.3_OPL  ESC_1.4_OPL  ESC_1.5_OPL  ESC_1.6_OPL ESC_1.10_OPL
   0.4131797    1.2472462    1.3175530    1.3269073    1.0525362    1.2882773

> head(gmpr.size.factor.1)
 ESC_1.1_OPL  ESC_1.3_OPL  ESC_1.4_OPL  ESC_1.5_OPL  ESC_1.6_OPL ESC_1.10_OPL
   0.3837041    1.4402244    1.4374717    1.5486711    1.2279985    1.3421698

By chance is one of these a later version than the other? Which of these would you recommend for use?

Thank you for your time!

Vince