For this, add a new genome "setup" to our subversion (like git) data repository.

Downloaded the Jun. 2020 GRCm39/mm39 mm39.2bit file from https://hgdownload.soe.ucsc.edu/goldenPath/mm39/bigZips/

Had to use ftp to get the gene2accession.gz file, there is some issue with trying to gunzip -c an ftp file, so I just pulled the whole file. I was able to download the Mus_musculus.gene_info file with no issues, but it is small.

ftp ftp://ftp.ncbi.nlm.nih.gov/gene/DATA/
get gene2accession.gz

NCBI:txid10090 for mouse:

gunzip -c gene2accession.gz | grep '^10090\t' > ~/Desktop/jiraIssues/3330/10090.gene2accession.txt

I pulled the refGene and ncbiRefSeq files as bed files. UCSC table browser.
I then created bed14 files for the refGene and ncbiRefSeq files.

ucscToBedDetail.py -a 10090.gene2accession.txt -g Mus_musculus.gene_info mm39_refGene.bed.gz M_musculus_Jun_2020_refGene.bed
ucscToBedDetail.py -a 10090.gene2accession.txt -g Mus_musculus.gene_info mm39_ncbiRefSeq.bed.gz M_musculus_Jun_2020_ncbiRefSeq.bed

I then sorted the two bed files and tabix indexed them. I was getting a lot of errors with tabix (version 1.17) where it was complaining about coordinates being zero and did I forget to add the -0 flag to indicate zero based coordinates. I don't remember this being an issue before, but since bed files are 0-based I added the flag.

sort -k1,1 -k2,2n M_musculus_Jun_2020_refGene.bed | bgzip > M_musculus_Jun_2020_refGene.bed.gz
tabix -0 -s 1 -b 2 -e 3 M_musculus_Jun_2020_refGene.bed.gz

sort -k1,1 -k2,2n M_musculus_Jun_2020_ncbiRefSeq.bed | bgzip > M_musculus_Jun_2020_ncbiRefSeq.bed.gz
tabix -0 -s 1 -b 2 -e 3 M_musculus_Jun_2020_ncbiRefSeq.bed.gz

The all_mRNA, and all_est files I pulled as psl files by selecting the output format as all fields from selected table. Once downloaded, I used the following code:

gunzip \-c M_musculus_Jun_2020_all_est.psl.gz | grep -v bin | cut -f2- > M_musculus_Jun_2020_all_est.psl
sort -k14,14 -k16,16n M_musculus_Jun_2020_all_est.psl > sorted.psl
mv sorted.psl M_musculus_Jun_2020_all_est.psl
bgzip M_musculus_Jun_2020_all_est.psl
tabix -s 14 -b 16 -0 M_musculus_Jun_2020_all_est.psl.gz

gunzip \-c M_musculus_Jun_2020_all_mrna.psl.gz | grep -v bin | cut -f2- > M_musculus_Jun_2020_all_mrna.psl
sort -k14,14 -k16,16n M_musculus_Jun_2020_all_mrna.psl > sorted.psl
mv sorted.psl M_musculus_Jun_2020_all_mrna.psl
bgzip M_musculus_Jun_2020_all_mrna.psl
tabix -s 14 -b 16 -0 M_musculus_Jun_2020_all_mrna.psl.gz

I have placed the DM6 quickload in CyVerse for testing.

To test:
In IGB,

1. Add https://data.cyverse.org/dav-anon/iplant/home/nowlanf/Mouse_2020/quickload as a new Data Source in IGB (IntegratedGenomeBrowser > Settings > Data Sources > Add...).
2. Select the Mus musculus Species and the M_musculus_Jun_2020 Genome Version.
3. The RefGene track should appear and should automatically load data.
4. Navigate to: chr1:36,405,677-36,408,789
5. Click Load Sequence.
6. Residues (ATCG) should load (may take a little while, CyVerse can be slow).
7. Check that there are no errors in the log.
8. Under Available Data in the Data Access tab, click the checkbox for mRNA, EST, and NCBI RefSeq
9. Click Load Data
10. Gene annotations for each track should load.
11. Check that there are no errors in the log.
12. Check that the UCSC (DAS) folder appears under Available Data in the Data Access tab.
13. Click the checkbox for ncbiRefSeqCurated.
14. Click Load Data.
15. Check that there are no errors in the log.

Other genomes that should be updated based on UCSC:
rat - IGBF-3362
chicken - IGBF-3361

I noticed that the psl files (EST and mRNA) have some kind of off by one error. The listed end position is +1 compared to where the annotations actually end (positive strand alignments). For example, if a mRNA on the positive strand says that it starts at 100 and ends at 150, in IGB the annotation will be drawn overlapping bases 100 - 149. If on the negative strand the IGB reported start is +1 to where the annotation is actually drawn. This indicates that the tEnd (https://genome.ucsc.edu/FAQ/FAQformat.html#format2) column is effectively +1 compared to what is actually drawn in IGB. Note that the annotations in IGB appear to be drawn in the correct location, i.e. intron exon boundaries line up correctly. The only issue is that the IGB reported start/end is off by one, depending on positive/negative strand.

This issue is present in older psl files (M_musculus_Dec_2011) and on older versions of IGB (tested 9.0.2).

As this issue is not new or specific to this ticket, I have created IGBF-3363 to address it.

Review: I was able to get through each step of the instructions and ran into no errors. The following is an image of my screen after finishing.

IGB_Mus-musculus.pdf

If the image is correct than the ticket is ready to move forward!

Notes from pushing changes to subversion repository: https://svn.bioviz.org/viewvc/genomes/quickload/M_musculus_Jun_2020/

Comment for subversion update
IGBF-3330: Add Mus musculus Jun 2020 genome;Jun 29, 2023

Need to update the following files:
contents.txt -> M_musculus_Jun_2020 Mus musculus (Jun 2020) mouse (GRCm39/mm39)
.htaccess -> AddDescription "Mus musculus (Jun 2020) mouse (GRCm39/mm39)" M_musculus_Jun_2020

Basic steps to checking in files and adding new rat folder with files:

1. Modify the contents.txt and .htaccess files
2. Drag the mouse folder into the quickload folder
   then:

   svn status
   svn ci -m "IGBF-3330: Add Mus musculus Jun 2020 genome;Jun 29, 2023" quickload/.htaccess
   svn ci -m "IGBF-3330: Add Mus musculus Jun 2020 genome;Jun 29, 2023" quickload/contents.txt
   svn add quickload/M_musculus_Jun_2020
   svn ci -m "IGBF-3330: Add Mus musculus Jun 2020 genome;Jun 29, 2023"
   

New mouse genome has been added to subversion repository.

Next step, deploy to quickload site.

Data are deployed. For details on where the data are deployed, and how to test, see
linked ticket titled "Add rn7 rat genome to IGB".

Tested on Mac on IGB release (9.1.10).
Able to load all data (RefGene, NCBI RefSeq, mRNA, EST, sequence file) on scidas and igbquickload.org

Closing ticket.