j-berg
commented
4 years ago Hi @cjdecker,
The GTF modification step was designed specifically with Ensembl formatting in mind. This error is arising due to a difference in formatting between UCSC and Ensembl GTF formatting, where Ensembl has a separate record summarizing the entire gene, labeled gene in column 3 of the GTF, whereas UCSC appears to exclude this record (at least by default). The geneInfo.tab file is output by STAR and is not involved in the GTF modification for CDS truncation, etc.
I think the most straightforward solution to this issue would be for me to include a UCSC flag during this step that can be provided and will help this module handle this difference in formatting. I should be able to get to this in the next few days and can at least get a sense of how tractable this solution will be. I will keep you updated.
Jordan
cjdecker
I ran the following command to curate reference for analyzing ribosome profiling data.
xpresspipe curateReference --output ./ --fasta genome_fasta/ --gtf ./transcripts.gtf --protein_coding --truncate --longest_transcript --sjdbOverhang 37
And I got this message before the process ended File "/Users/carolyndecker/anaconda2/envs/xpresspipe/lib/python3.7/site-packages/XPRESSpipe-0.6.0-py3.7.egg/xpresspipe/gtfModify.py", line 244, in get_chunks raise Exception('No gene or transcript records found in GTF') Exception: No gene or transcript records found in GTF
I noticed in the geneInfo.tab file that it says "MissingGeneType" $ head -3 geneInfo.tab 70481 NM_001276351 NM_001276351 MissingGeneType NR_075077 NR_075077 MissingGeneType
I am collaborating with a group on this project and they use hg38_refseq gtf file rather than ensembl gtf. I am wondering if this is an issue with XPRESSpipe not being able to parse required information from the refseq gtf (see below example lines from transcripts.gtf file).
Would I be able to get around this issue by modifying the geneInfo.tab file so that transcripts with NM prefix are assigned as protein coding and NR non coding?
Appreciate your help in resolving this issue, Carolyn $ head transcripts.gtf chr1 hg38_refGene stop_codon 67093005 67093007 0.000000 - . gene_id "NM_001276351"; transcript_id "NM_001276351"; chr1 hg38_refGene CDS 67093008 67093604 0.000000 - 0 gene_id "NM_001276351"; transcript_id "NM_001276351"; chr1 hg38_refGene exon 67092166 67093604 0.000000 - . gene_id "NM_001276351"; transcript_id "NM_001276351"; chr1 hg38_refGene CDS 67095235 67095421 0.000000 - 1 gene_id "NM_001276351"; transcript_id "NM_001276351"; chr1 hg38_refGene exon 67095235 67095421 0.000000 - . gene_id "NM_001276351"; transcript_id "NM_001276351"; chr1 hg38_refGene CDS 67096252 67096321 0.000000 - 2 gene_id "NM_001276351"; transcript_id "NM_001276351"; chr1 hg38_refGene exon 67096252 67096321 0.000000 - . gene_id "NM_001276351"; transcript_id "NM_001276351"; chr1 hg38_refGene CDS 67115352 67115464 0.000000 - 1 gene_id "NM_001276351"; transcript_id "NM_001276351"; chr1 hg38_refGene exon 67115352 67115464 0.000000 - . gene_id "NM_001276351"; transcript_id "NM_001276351"; chr1 hg38_refGene CDS 67125752 67125909 0.000000 - 0 gene_id "NM_001276351"; transcript_id "NM_001276351";