TU_filter errors

[YuH-LIU]

Hi
here is the error I got after I ran the app:
Listening on http://127.0.0.1:3061/
Warning: Error in [[<-: 1 elements in value to replace 0 elements
88: stop
87: [[<-
85: $<-
83: TU_prep [/home/s2868206/TU/TU_filter/app.R#1866]
82: bin_transcribed_TU [/home/s2868206/TU/TU_filter/app.R#1697]
81: observe [/home/s2868206/TU/TU_filter/app.R#647]
80: observer:observeEvent(input$go)
1: runApp

I tried to input the gtf with and without 'chr' in the column of chromosome but I still got the same error.
May I ask what I can do to fix it? thanks

[shaorray]

Hi,

This issue seems to relate with the assignment to GenomicRanges objects.

Could you check in R console, if assignment works?

For example:

library(GenomicRanges)

gr = GRanges(
    seqnames = Rle(c("chr1", "chr2", "chr1", "chr3"), c(1, 3, 2, 4)),
    ranges = IRanges(101:110, end = 111:120, names = head(letters, 10)),
    strand = Rle(strand(c("-", "+", "*", "+", "-")), c(1, 2, 2, 3, 2)),
    score = 1:10,
    GC = seq(1, 0, length=10))

gr$expr = 0   # check if this line works?

[YuH-LIU]

yes! that works!

Would it have anything to do with which aligner I used (STAR) or my input bam files.
My data is TT-seq and I did alignment with chimeric genome (mouse and drosophila). Hope those info are helpful to figure out what is going on there! thanks!

[shaorray]

After interrupting the run, is there a TU.gr object in global environment? If yes, is its length = 0? How about the length of TU_input ?

Gene reference (ensembl or GENCODE) is better to be the same file as STAR index, and is ok to be chimeric.

[YuH-LIU]

I didn't get TU.gr.
here is how TU_input looks like:

TU_input
GRanges object with 19917 ranges and 5 metadata columns:
seqnames ranges strand | id gene_id exonOV type gene_type
|
[1] chr1 4767001-4785801 + | 1 FALSE ncRNA
[2] chr1 4800201-4802801 + | 2 FALSE ncRNA
[3] chr1 4806801-4807401 + | 3 FALSE ncRNA
[4] chr1 4856401-4857601 + | 4 FALSE ncRNA
[5] chr1 4879601-4880601 + | 5 FALSE ncRNA
... ... ... ... . ... ... ... ... ...
[19913] chrY 47714201-47720601 - | 19913 FALSE ncRNA
[19914] chrY 90717401-90743401 - | 19914 FALSE ncRNA
[19915] chrY 90785201-90786401 - | 19915 FALSE ncRNA
[19916] chrY 90790401-90815201 - | 19916 FALSE ncRNA
[19917] chrY 90819401-90821201 - | 19917 FALSE ncRNA

---

seqinfo: 21 sequences from an unspecified genome; no seqlengths

Yes, I used the same one as for STAR index.

[shaorray]

The process seems to stop at TU cutoff step. Could you check if "TU_input$gene_type" has any "protein_coding" attribute?

If yes, your global environment has a "gene.gr" object, right?