Regarding the issue of generating the final version of the GFF format file

[jihui-boop]

Dear Professor, I would like to inquire how the "gene_symbols_edited.tsv" file was generated. This is because I ran the "PostprocessReformatFinalGFF.R" script in the "6-AfterAnnotation" section.
I haven't found this file for this script. The teacher hopes that you can spare some time to give me guidance.

[jihui-boop]

Dear Professor, I have now obtained the content of full_annotation.geneSymbols.gtf as follows
chr1 Mikado_loci_transcript 22240 23458 . - . transcript_id "mikado.chr1G2.1"; gene_id "mikado.chr1G2.1"; Name "mikado.chr1G2"; CDS_Name "mikado.chr1G2.1.CDS1"
chr1 Mikado_loci exon 22240 22784 . - . transcript_id "mikado.chr1G2.1"; Name "mikado.chr1G2.1.five_prime_UTR1";
chr1 Mikado_loci exon 23363 23458 . - . transcript_id "mikado.chr1G2.1"; Name "mikado.chr1G2.1.five_prime_UTR2";
chr1 Mikado_loci CDS 22389 22478 . - 0 transcript_id "mikado.chr1G2.1";
chr1 Mikado_loci/transcript 23629 30961 . + . transcript_id "mikado.chr1G1.1"; gene_id "mikado.chr1G1.1"; Name "XP_003127075"
chr1 Mikado_loci exon 23629 23911 . + . transcript_id "mikado.chr1G1.1"; Name "mikado.chr1G1.1.exon1";
chr1 Mikado_loci exon 25017 25259 . + . transcript_id "mikado.chr1G1.1"; Name "mikado.chr1G1.1.exon2";
chr1 Mikado_loci exon 25579 25710 . + . transcript_id "mikado.chr1G1.1"; Name "mikado.chr1G1.1.exon3";
However, I couldn't find the file gene_symbols_edited.tsv.

[DustinSokolowski]

Hey Jihui,

First of all I'm very happy that you were able to work through the tutorial. The file is the output of the "EditGeneSymbolTable.Rmd" markdown, which was formerly named "gene_symbols_noCopies.edited.tsv". This was an error on our end, so thank you for catching this, it is now fixed.

I have also been making minor improvements to the PostprocessReformatFinalGFF.R script as I use GenAnT to annotate various mammalian genomes for other projects. I was going to make a larger commit later with some added material (e.g., pseudogenes, improved rRNA mapping) but I have updated it now as it may be helpful for you. The two changes are.

1. if two genes are given the same gene symbol, ABC-copy1 and ABC-copy2 are now ordered by gene quality score as described by mikado, rather than simply which gene has an earlier "chr:start" co-ordinate on the genome.

2. The final outputs include a gene_symbols_metrics.tsv file, which combines the gene symbol table and the mikado_metrics.tsv output file. I have found this quite valuable for downstream work (e.g., filtering gff files to improve 1-1 ortholog mapping for comparative genomics).

Nothing in how to run any of the scripts or pipeline changes, there's just an extra file now.

Thanks again for your interest!

[jihui-boop]

Thank you, Professor. Regarding the above issue, after your guidance, it has been resolved. However, the content of the full_annotation.geneSymbols.gtf that I generated now is: hr1 Mikado_loci_transcript 22240 23458 . - . transcript_id "mikado.chr1G2.1"; gene_id "mikado.chr1G2.1"; Name "mikado.chr1G2"; CDS_Name "mikado.chr1G2.1.CDS1"
chr1 Mikado_loci exon 22240 22784 . - . transcript_id "mikado.chr1G2.1"; Name "mikado.chr1G2.1.five_prime_UTR1";
chr1 Mikado_loci exon 23363 23458 . - . transcript_id "mikado.chr1G2.1"; Name "mikado.chr1G2.1.five_prime_UTR2";
chr1 Mikado_loci CDS 22389 22478 . - 0 transcript_id "mikado.chr1G2.1";
chr1 Mikado_loci/transcript 23629 30961 . + . transcript_id "mikado.chr1G1.1"; gene_id "mikado.chr1G1.1"; Name "XP_003127075"
chr1 Mikado_loci exon 23629 23911 . + . transcript_id "mikado.chr1G1.1"; Name "mikado.chr1G1.1.exon1";
chr1 Mikado_loci exon 25017 25259 . + . transcript_id "mikado.chr1G1.1"; Name "mikado.chr1G1.1.exon2";
chr1 Mikado_loci exon 25579 25710 . + . transcript_id "mikado.chr1G1.1"; Name "mikado.chr1G1.1.exon3";
chr1 Mikado_loci exon 26460 26563 . + . transcript_id "mikado.chr1G1.1"; Name "mikado.chr1G1.1.exon4";
chr1 Mikado_loci exon 28020 28133 . + . transcript_id "mikado.chr1G1.1"; Name "mikado.chr1G1.1.exon5";
chr1 Mikado_loci exon 29265 30961 . + . transcript_id "mikado.chr1G1.1"; Name "mikado.chr1G1.1.three_prime_UTR1";
chr1 Mikado_loci CDS 23629 23911 . + 0 transcript_id "mikado.chr1G1.1"; Name "mikado.chr1G1.1.CDS1";
How can I know the actual name of the corresponding gene? I need to know the function of the corresponding gene. Teacher, I hope you can spare some time to guide me.

[DustinSokolowski]

Hey Jihui,

The GFF looks good! I can tell this because of this line:

chr1 Mikado_loci/transcript 23629 30961 . + . transcript_id "mikado.chr1G1.1"; gene_id "mikado.chr1G1.1"; Name "XP_003127075"

XP_003127075 is not anything that is made by mikado, but is a gene transferred from your reference species based on https://www.genecards.org/card/PDCD2L). The problem is that rather than annotating with a unique gene symbol from your reference species, it is annotating with the protein_id.

I think that what you'll have to tinker with is in the EditGeneSymbolTable markdown. https://github.com/BaderLab/GenAnT/blob/main/6-AfterAnnotation/EditGeneSymbolTable.Rmd

When it runs the "swap_symbols" function: swap_symbols <- function(gene_symbols , symbol_colname, gene_key, matching_column = "transcriptID" ), I'd check the matching_column parameter. My guess is that the column name used in your key between gene symbol and gene ID does not match what you put into "matching_column". Given you have XP_003127075, it'll probably have to be proteinID or protein_id or something.

Sorry about not really having a way to standardize this, it's why we chose to make EditGeneSymbolTable a markdown rather than just an R script people can run, since it's dependent on how the reference species are formatted.

Best,
Dustin