Question

error with detectTranscripts

0

Entering edit mode

warrena • 0

@warrena-16186

Last seen 6.3 years ago

Greetings,

I am attempting to use detectTranscripts() from groHMM using bams from a PRO-seq experiment. I am getting the following error:

Error in Fp[[i]] + 1 : non-numeric argument to binary operator In addition: Warning messages: 1: In mclapply(readsList, function(x) { : all scheduled cores encountered errors in user code 2: In mclapply(readsList, function(x) { : all scheduled cores encountered errors in user code

Here is my code:

s1 <- as(readGAlignments("poolA_index1_sample.bam"), "GRanges")
s2 <- as(readGAlignments("poolA_index2_sample.bam"), "GRanges")
S0m <- c(s1, s2)
Sall <- sort(c(S0m))
hmmResult <- detectTranscripts(Sall, LtProbB=-200, UTS=5, threshold=1)

From what I can tell, my data are formatted identically to the data from the groHMM vignette (those data give the expected output), e.g.,

> head(Sall)
GRanges object with 6 ranges and 0 metadata columns:
      seqnames             ranges strand
         <Rle>          <IRanges>  <Rle>
  [1]     chr1 [3000917, 3000935]      +
  [2]     chr1 [3010291, 3010321]      +
  [3]     chr1 [3102785, 3102808]      +
  [4]     chr1 [3190350, 3190371]      +
  [5]     chr1 [3230646, 3230667]      +
  [6]     chr1 [3282720, 3282739]      +
  -------
  seqinfo: 66 sequences from an unspecified genome

I have also verified that my genomic ranges contain all numeric entries, e.g.:

rang <- as.data.frame( ranges(Sall) ) inds = which(is.numeric(rang[,1])==F) # integer(0)

Assistance or advice would be greatly appreciated.

groHMM • 1.1k views

ADD COMMENT • link updated 6.4 years ago by lee.kraus • 0 • written 6.4 years ago by warrena • 0

0

Entering edit mode

Hello! Removing the random chr annotations (chrUn_gl0000230, chr9_gl000198_random) from your bam file should take of that error with the current version of groHMM package. Please try:

Sall_stdchr <- keepStandardChromosomes(Sall)
hmmResult <- detectTranscripts(Sall_stdchr, LtProbB=-200, UTS=5, threshold=1)

Best,

Anusha

ADD REPLY • link 6.4 years ago anusha.nagari ▴ 50

0

Entering edit mode

Thanks, Anusha. This worked. I'll add that for users of R 3.4.4 I had to call keepStandardChromosomes() with pruning.mode="coarse". Here is the code:

require(groHMM)
library(IRanges)
library(GenomicRanges)
library(Rsamtools)
library(GenomicAlignments)

data <- readGAlignments("poolA_index1_sample.bam")
data_gr <- granges(data)
data_gr <- sort(data_gr)
gr <- keepStandardChromosomes(data_gr, pruning.mode="coarse")
hmmResult <- detectTranscripts(gr, LtProbB=-200, UTS=5, threshold=1)

ADD REPLY • link 6.4 years ago warrena • 0

score 0 · Answer 1 · 2018-07-09

Hi Anusha, Can you please respond to this? Thanks, Lee W. Lee Kraus 607-227-8726 Lee.Kraus@utsouthwestern.edu<mailto:lee.kraus@utsouthwestern.edu> On Jul 9, 2018, at 1:31 PM, warrena [bioc] <noreply@bioconductor.org<mailto:noreply@bioconductor.org>> wrote: Activity on a post you are following on support.bioconductor.org<https: support.bioconductor.org=""> User warrena<https: support.bioconductor.org="" u="" 16186=""/> wrote Question: error with detectTranscripts<https: support.bioconductor.org="" p="" 110852=""/>: Greetings, I am attempting to use detectTranscripts() from groHMM using bams from a PRO-seq experiment. I am getting the following error: Error in Fp[[i]] + 1 : non-numeric argument to binary operator In addition: Warning messages: 1: In mclapply(readsList, function(x) { : all scheduled cores encountered errors in user code 2: In mclapply(readsList, function(x) { : all scheduled cores encountered errors in user code Here is my code: s1 <- as(readGAlignments("poolA_index1_sample.bam"), "GRanges") s2 <- as(readGAlignments("poolA_index2_sample.bam"), "GRanges") S0m <- c(s1, s2) Sall <- sort(c(S0m)) hmmResult <- detectTranscripts(Sall, LtProbB=-200, UTS=5, threshold=1) From what I can tell, my data are formatted identically to the data from the groHMM vignette (those data give the expected output), e.g., > head(Sall) GRanges object with 6 ranges and 0 metadata columns: seqnames ranges strand <rle> <iranges> <rle> [1] chr1 [3000917, 3000935] + [2] chr1 [3010291, 3010321] + [3] chr1 [3102785, 3102808] + [4] chr1 [3190350, 3190371] + [5] chr1 [3230646, 3230667] + [6] chr1 [3282720, 3282739] + ------- seqinfo: 66 sequences from an unspecified genome I have also verified that my genomic ranges contain all numeric entries, e.g.: rang <- as.data.frame( ranges(Sall) ) inds = which(is.numeric(rang[,1])==F) # integer(0) Assistance or advice would be greatly appreciated. ________________________________ Post tags: groHMM You may reply via email or visit error with detectTranscripts ________________________________ UT Southwestern Medical Center The future of medicine, today.