This package provides thermal stable RNA 2ndary structures predicted on full length transcripts mm10 and hg19. The predicted structures are used to filter the bisulfite sequening sites that have potentially incomplete bisulfite conversion on them.
- First, install the package in R
devtools::install_github("ZW-xjtlu/rBS2ndStructure")
library(rBS2ndStructure)- Then, you could use the following data containing the ranges of the structures provided by the package. Use
?to check their documentations.
library(GenomicRanges)
?rBS2ndStructure::Struc_hg19
?rBS2ndStructure::Struc_mm10- You could directly use the provided GRanges object to filter your sites.
rBS_gr = GRanges(seqnames = rBS_df$`#SeqID`, strand = rBS_df$refStrand, ranges = IRanges(start = rBS_df$refPos, width = 1))
rBS_gr_filtered <- rBS_gr[!rBS_gr %over% Struc_hg19]- Alternatively, you could try to filter the sites overlapped with RNA secondary structures with the function
rBS_2ndStructure_Filter, the function also helps you mask all the sites on introns and intergenic sequences.
rBS_gr = GRanges(seqnames = rBS_df$`#SeqID`, strand = rBS_df$refStrand, ranges = IRanges(start = rBS_df$refPos, width = 1))
rBS_gr$mcols = p.adjust(rBS_df$`p-value_mState`,method = "BH") < .05
rBS_gr_filtered <- rBS_2ndStructure_Filter(rBS_gr,"hg19")- Finally, you could try to predict RNA 2ndary structures of your own transcriptome with RNAfold of viennaRNA package installed on your system.
#Extract transcript sequences in fasta files from BSgenome and txdb
library(BSgenome.Mmusculus.UCSC.mm10)
library(TxDb.Mmusculus.UCSC.mm10.knownGene)
txdb <- TxDb.Mmusculus.UCSC.mm10.knownGene
tx_seq_extraction(MMusculus,txdb,getwd())
#Run RNAfold in multiple computational clusters
RNAfold(1:30,"Small","Fsm")
RNAfold(1:30,"Middle","Fmd")
RNAfold(1:30,"Large","Flg")
# Assemble the predicted transcripts
rfold_assembly_tx(txdb,RfdNames = c("Fsm","Fmd","Flg"))- Notice that you could run
RNAfoldon command line by yourself, the functionRNAfoldonly works if you implementqsubcommand on your computer system.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent