baltica with human genome from GENECODE

[sailorjupiter87]

Hi everyone,

I am new to the forum and I am still getting familiar with using Baltica (Ubuntu 22.04.5 LTS, Python 3.7.12, baltica 1.2.4). I encountered an issue during the final stages of analysis with my RNA-seq data (second strand library prep) aligned with STAR using the human genome GENECODE GRCh38.p4 release47 primary assembly (gencode.v47.primary_assembly.annotation.gtf and GRCh38.primary_assembly.genome.fa). I was able to obtain the outputs from the DJU method and the StringTie output, but at this point, baltica stopped with the following error message:

"Activating singularity image /home/cappelli/.baltica/singularity/8c8a3574e99286bac514a846a13eea14.simg
Processing DJU method output
Processing de novo annotation
Warning message:
In .Seqinfo.mergexy(x, y) :
Each of the 2 combined objects has sequence levels not in the other:

• in 'x': KI270438.1, KI270512.1, KI270709.1, KI270729.1, KI270732.1, KI270747.1
• in 'y': GL000216.2, KI270710.1, KI270713.1, KI270714.1, KI270719.1, KI270720.1, KI270726.1, KI270746.1, KI270749.1, KI270753.1, KI270755.1
  Make sure to always combine/compare objects based on the same reference
  genome (use suppressWarnings() to suppress this warning).
  Preparing annotation
  Proceding with integration
  Error in validObject(x) : invalid class “IRanges” object:
  'width(x)' cannot contain negative integers
  Calls: lapply ... end<- -> end<- -> end<- -> .set_IRanges_end -> validObject
  In addition: Warning message:
  In .Seqinfo.mergexy(x, y) :
  Each of the 2 combined objects has sequence levels not in the other:
• in 'x': chr1, chr2, chr3, chr4, chr5, chr6, chr7, chr8, chr9, chr10, chr11, chr12, chr13, chr14, chr15, chr16, chr17, chr18, chr19, chr20, chr21, chr22, chrX, chrY, chrM, GL000008.2, GL000009.2, GL000195.1, GL000205.2, GL000213.1, GL000214.1, GL000216.2, GL000218.1, GL000219.1, GL000221.1, GL000224.1, GL000225.1, KI270442.1, KI270706.1, KI270710.1, KI270711.1, KI270712.1, KI270713.1, KI270714.1, KI270717.1, KI270718.1, KI270719.1, KI270720.1, KI270722.1, KI270726.1, KI270727.1, KI270728.1, KI270731.1, KI270734.1, KI270741.1, KI270742.1, KI270743.1, KI270744.1, KI270745.1, KI270746.1, KI270748.1, KI270749.1, KI270750.1, KI270751.1, KI270753.1, KI270755.1
• in 'y': 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, X, Y, M
  Make sure to always combine/compare objects based on the same reference
  genome (use suppressWarnings() to suppress this warning).
  Execution halted"

Upon reopening the outputs (junction.csv), I noticed that in some files the chromosomes were labeled with just numbers, while in others they had the "chr" prefix. I standardized the chromosome names by adding the "chr" prefix to all chromosomes in each file and ran Baltica analysis. However, I encountered the same error again. Since the error also referred to StringTie, I checked the folder generated by StringTie to look at the merged.combined.gtf file and compared it with the Baltica analysis run on the same RNA-seq data, but aligned to a human genome downloaded from Ensembl (it is hg38, but I don't know the release because it was provided by the company that sequenced my samples). Essentially, these two merged.combined.gtf files differ in the presence of additional columns and also in the inclusion of "scaffolds" in the file generated from the genome downloaded from GENECODE.
How can I solve this problem? Is it possible to use a genome downloaded from GENECODE to perform the assembly for use in Baltica?

Many thanks in advance,

Sara