search

CRI-iAtlas / ImmuneSubtypeClassifier

An R package for classification of immune subtypes, in cancer, using gene expression data.
Other
40 stars 23 forks source link

Data preparation of ImmuneSubtypeClassifier Package #3

Open huangwb8 opened 4 years ago

huangwb8 commented 4 years ago

Hello~ After some tests about this package, I think it's easy to use. Meanwhile, the capacity of concordance is robust, according to the reported result.Here are my questions:

1.Are there any limitations about the form of gene expression data? For example,is it available if I use gene expression data from Affymetrix miroarray? Or it's only fitted in RPKM data?

2.Before go to the pipeline, shall gene matrix be normalized or scaled? Thanks~

Tim0thy1 commented 4 years ago

Hello~ After some tests about this package, I think it's easy to use. Meanwhile, the capacity of concordance is robust, according to the reported result.Here are my questions:

1.Are there any limitations about the form of gene expression data? For example,is it available if I use gene expression data from Affymetrix miroarray? Or it's only fitted in RPKM data?

2.Before go to the pipeline, shall gene matrix be normalized or scaled? Thanks~

Hi, huangwb8

Are you successfully install the R package? It look like 'ImmuneSubtypeClassifier ' is not available for R 3.6 . What version of R you running?

install_github("Gibbsdavidl/ImmuneSubtypeClassifier") Downloading GitHub repo Gibbsdavidl/ImmuneSubtypeClassifier@master √ checking for file 'C:\Users***\AppData\Local\Temp\RtmpOMQkof\remotes54b419a06345\Gibbsdavidl-ImmuneSubtypeClassifier-30e6215/DESCRIPTION'

  • preparing 'ImmuneSubtypeClassifier': √ checking DESCRIPTION meta-information ...
  • checking for LF line-endings in source and make files and shell scripts
  • checking for empty or unneeded directories
  • looking to see if a 'data/datalist' file should be added NB: this package now depends on R (>= 3.5.0)
huangwb8 commented 4 years ago

Yes! I've installed the package. Thank you very much!

发自我的iPhone

------------------ Original ------------------ From: Tim0thy1 <notifications@github.com> Date: Wed,Sep 4,2019 3:54 PM To: Gibbsdavidl/ImmuneSubtypeClassifier <ImmuneSubtypeClassifier@noreply.github.com> Cc: Weibin Huang <654751191@qq.com>, Author <author@noreply.github.com> Subject: Re: [Gibbsdavidl/ImmuneSubtypeClassifier] Data preparation of ImmuneSubtypeClassifier Package (#3)

Hi, huangwb8

Are you successfully install the R package? It look like 'ImmuneSubtypeClassifier ' is not available for R 3.6 . What version of R you running?

install_github("Gibbsdavidl/ImmuneSubtypeClassifier") Downloading GitHub repo Gibbsdavidl/ImmuneSubtypeClassifier@master √ checking for file 'C:\Users***\AppData\Local\Temp\RtmpOMQkof\remotes54b419a06345\Gibbsdavidl-ImmuneSubtypeClassifier-30e6215/DESCRIPTION'

preparing 'ImmuneSubtypeClassifier': √ checking DESCRIPTION meta-information ...

checking for LF line-endings in source and make files and shell scripts

checking for empty or unneeded directories

looking to see if a 'data/datalist' file should be added NB: this package now depends on R (>= 3.5.0)

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub, or mute the thread.

Tim0thy1 commented 4 years ago

@huangwb8 I am not author. I have a problem about install.

Gibbsdavidl commented 4 years ago

Thank you for trying it out!

Hello~ After some tests about this package, I think it's easy to use. Meanwhile, the capacity of concordance is robust, according to the reported result.Here are my questions:

1.Are there any limitations about the form of gene expression data? For example,is it available if I use gene expression data from Affymetrix miroarray? Or it's only fitted in RPKM data?

I believe microarray data should be OK. Maybe! I've never tried it. But... see my next answer.

2.Before go to the pipeline, shall gene matrix be normalized or scaled?

Please don't scale or normalize the genes across samples, that's the only rule.

The classifier does not depend on the value of gene expression... it works on gene pairs (or.. same idea.. but gene-set-pairs). So, for a given sample, if gene_A > gene_B then the feature value is 1, otherwise 0 (for this sample). But, if we normalize genes by ... say median scaling the genes across samples... then the relationship between our gene pairs changes, and damages the prediction.

I hope that helps. -dave

Thanks~

Gibbsdavidl commented 4 years ago

@huangwb8 I am not author. I have a problem about install.

Hi there, could you make a new issue, and we can get it sorted?

Thanks! -dave

huangwb8 commented 4 years ago

i'm sorry~my computer OP is WIN10, and I use 3.5 R version. my suggestion is: you can try a local installation. Download the package as *.tar.gz, then install it via Rstudio or R GUI.