Question

Bioconductor packages for machine learning with RNAseq

1

Entering edit mode

jonessara770 ▴ 10

@jonessara770-11760

Last seen 6.7 years ago

Hello all,

can you please give a list of packages that have been developed to apply machine learning models on RNAseq data?

Thanks

Sara

bioconductor • 1.1k views

ADD COMMENT • link updated 8.1 years ago by Wolfgang Huber ★ 13k • written 8.1 years ago by jonessara770 ▴ 10

0

Entering edit mode

There is a bioconductor package "MLInterfaces" that contains interfaces to R machine learning procedures such as svm, random forest, knn, and etc. Those procedures are built for general purpose, yet you might find it quite helpful for expression data.

ADD REPLY • link 8.1 years ago Chao-Jen Wong ▴ 40

score 1 · Answer 1 · 2016-11-03

1

Entering edit mode

James W. MacDonald 67k

@james-w-macdonald-5106

Last seen 8 hours ago

United States

If you go to the package listing page, on the left hand side are a set of 'biocViews', one of which is StatisticalMethod. Machine learning is a broad category that encompasses several of the methods there, but you should be able to look through the choices and see which package(s) are applicable.

ADD COMMENT • link 8.1 years ago James W. MacDonald 67k

score 1 · Answer 2 · 2016-11-03

Sara

perhaps you are referring to characteristics of RNA-Seq data that make it necessary to "preprocess" the data before applying regular ML methods: normalisation for library size and transformation to make the data less skewed and less heteroskedastic. Have a look at the variance stabilizing transformation in DESeq2. It's also described in the vignette.

Hope this helps

Wolfgang