salmon2deseq: Iniitialize a DESeq2 object from salmon output.

In Tong-Chen/YSX: For Yishengxin Training

Description

Iniitialize a DESeq2 object from salmon output.

Usage

salmon2deseq(
  salmon_file_list,
  sampleFile,
  design,
  covariate = NULL,
  tx2gene = NULL,
  filter = NULL,
  rundeseq = T
)

Arguments

salmon_file_list	A two-column file with the first column as sample names and second column containing the path of quant.sf generated by salmon. Header line is required but column names do not matter. Below is an example (pay attention to the path of quant.sf) Samp path_quant.sf untrt_N61311 untrt_N61311/untrt_N61311.salmon.count/quant.sf untrt_N052611 untrt_N052611/untrt_N052611.salmon.count/quant.sf untrt_N080611 untrt_N080611/untrt_N080611.salmon.count/quant.sf untrt_N061011 untrt_N061011/untrt_N061011.salmon.count/quant.sf trt_N61311 trt_N61311/trt_N61311.salmon.count/quant.sf trt_N052611 trt_N052611/trt_N052611.salmon.count/quant.sf trt_N080611 trt_N080611/trt_N080611.salmon.count/quant.sf trt_N061011 trt_N061011/trt_N061011.salmon.count/quant.sf
sampleFile	A file containing at least two columns. The first column is sample name just like the first column of salmon_file_list. Other columns are sample attributes. Normally one of sample attributes should contain the group information each sample belongs to. One simple example (conditions represent group information) Samp conditions untrt_N61311 untrt untrt_N052611 untrt untrt_N080611 untrt untrt_N061011 untrt trt_N61311 trt trt_N052611 trt trt_N080611 trt trt_N061011 trt Another example (3rd column meaning samples from two batches) Samp conditions batch untrt_N61311 untrt A untrt_N052611 untrt A untrt_N080611 untrt B untrt_N061011 untrt B trt_N61311 trt A trt_N052611 trt A trt_N080611 trt B trt_N061011 trt B
design	A column name from "sampleFile" like "conditions" in example. This will be used as group variable for DE tests. Currently only simple design is allowed. If one wants to model multiple variables, construct one representation of super variable as indicated in https://support.bioconductor.org/p/67600/#67612 may be useful.
covariate	Names of columns containing informations maybe covariates like batch effects or other sample info. Multiple covariates should be supplied as a vector.
tx2gene	Optional and only used if one want to get gene expression instead of transcript expression. A two-column file with the first column as transcript names and second column as gene names. Header line is required but column names do not matter. Below is an example of file contents. txname gene ENST00000456328 ENSG00000223972 ENST00000450305 ENSG00000223972 ENST00000488147 ENSG00000227232 ENST00000619216 ENSG00000278267 ENST00000473358 ENSG00000243485
filter	Filter genes with low read counts. Default genes with total reads count lower than half of number of samples will be filtered out. One can give any number here. Normally default is OK. The DESeq2 will ao auto filter too. Check https://www.bioconductor.org/packages/devel/bioc/vignettes/DESeq2/inst/doc/DESeq2.html.
rundeseq	Default TRUE. The function will perfrom deseq analysis using DESeq and return analyzed DESeqDataSet object. If FALSE, just return a DESeqDataSet object and one can run DESeqon it with more customed parameters.

Value

A DESeqDataSet object.

Examples

dds <- salmon2deseq(salmon_file_list, sampleFile, "conditions")
dds <- salmon2deseq(salmon_file_list, tx2gene=tx2gene,
                    sampleFile, "conditions")

Tong-Chen/YSX documentation built on Jan. 25, 2021, 2:49 a.m.