From: Colin <col...@gm...> - 2022-05-26 18:00:58
|
The code that is crashing is probably this, seeing renderDatasetSelector in the stack trace https://github.com/GMOD/jbrowse/blob/fd7ebb8a0802db9e09ac34c70689e7a9cfccc615/src/JBrowse/Browser.js#L1214-L1216 Since it is crashing there, I think maybe your manually configured datasets are not matching the Apollo configuration of the datasets. If you make them match up, it might fix it, and to do that, you may be able to make the dataset ID be the organism ID that Apollo gives. Would have to update both the datasets and the dataset_id This is where Apollo sets the dataset_id, uses organism ID https://github.com/GMOD/Apollo/blob/1728c280f9a18bd22e07f8b4b9f00a8e24432358/grails-app/controllers/org/bbop/apollo/JbrowseController.groovy#L435-L440 To be clear, the dataset ID is what is after the . in the conf format e.g. [datasets.make_this_match_the_organismID] url = ?data=data/nipponbare name = Nipponbare Also, the current generation of Apollo (Apollo 2) does not work with JBrowse 2, but we are working on a new Apollo version that will be integrated with JBrowse 2. It's not ready for use yet but you can keep an eye on https://github.com/GMOD/Apollo3 if interested -Colin On Thu, May 26, 2022 at 9:56 AM Justin Elser <jus...@or...> wrote: > I'm attaching the output from the chrome console. I don't think it's the > non-minified version of the JS though. How would I get that with the docker > version of apollo? > > Also, I've been asked to ask if Apollo works with JBrowse2? Would it be > better to use that instead of JBrowse1? > > Thank you for taking the time to help me. > > Justin > > On 5/25/22 4:51 PM, Colin wrote: > > [This email originated from outside of OSU. Use caution with links and > attachments.] > The logs that are probably most helpful are the "JavaScript console" or > developer console from chrome or firefox e.g. > https://balsamiq.com/support/faqs/browserconsole/ > > On Wed, May 25, 2022, 4:33 PM Justin Elser <jus...@or...> > wrote: > >> We may just use Apollo to serve both, but thought I might give it a try >> to see about JBrowse. >> >> I am attaching the output I get when I run the docker container with "-e >> WEBAPOLLO_DEBUG=true". >> >> The blank lines are in between it working and not working. IE, lines >> 1-483 are the initial startup and navigating to the site with Apollo >> working as expected. Lines 487 on are after changing >> data/nipponbare/tracks.conf to have: >> [general] >> dataset_id = nipponbare >> >> I do notice that after the change it is trying to open >> /data/temporary/apollo_data/34-Nipponbare/sample_data/json/volvox/successfully_run >> (line 579) where it isn't before. I don't understand why it would be >> looking for the sample data at all. I never loaded it as a track or >> anything. >> >> I also notice for the initial load where it works that it is complaining >> about the mime_type for .bam files. I did alter the .htaccess for the >> jbrowse but not apollo, so I think those are fixable and unrelated. Weirdly >> enough, the bam track does load fine in Apollo where it didn't in JBrowse >> until I made the change to .htaccess. >> >> If you meant some other logs, please let me know how to get those and I >> will send them to you. >> >> Thanks, >> Justin >> >> On 5/25/22 2:18 PM, Colin wrote: >> >> [This email originated from outside of OSU. Use caution with links and >> attachments.] >> One option is to use Apollo to serve your jbrowse instances, Apollo can >> serve public jbrowse instances so you may not need to configure jbrowse >> separately >> >> >> If you do want it separate, then we would probably need to debug that >> error. If you have the stack trace from the developer console, it might >> narrow down exactly where the error comes from. If the stack trace is >> minified and hard to read, can possibly use a "dev" version of jbrowse if >> it is too unreadable since it does not minify the builds! >> >> >> >> -Colin >> >> On Wed, May 25, 2022 at 1:03 PM Justin Elser < >> jus...@or...> wrote: >> >>> Yeah, so I guess the problem is that if I don't add the "dataset_id" to >>> each data/(species)/tracks.conf file, apollo works fine, but then I can't >>> get the dataset selector in jbrowse. As I have it right now, Apollo is >>> fine, but I have no way to switch to a different dataset in JBrowse. >>> >>> >>> I was messing around with jbrowse.conf, data/tracks.conf, and >>> data/(species)/tracks.conf, and was able to get a list of the species under >>> the "Genome" tab. But if I don't have anything in >>> data/(species)/tracks.conf, I don't. However, if I do add it like in the >>> documentation, the species are listed under there: >>> >>> But then, apollo breaks: >>> >>> >>> So, I guess I need to get the dataset_selector working in JBrowse, while >>> not breaking Apollo. >>> >>> I hope that clears it up more. >>> >>> Thanks, >>> Justin >>> >>> On 5/25/22 11:50 AM, Colin wrote: >>> >>> [This email originated from outside of OSU. Use caution with links and >>> attachments.] >>> I haven't evaluated the whole post here, but note that Apollo >>> automatically creates a dataset selector from assemblies that you configure >>> through that Apollo interface >>> >>> >>> -Colin >>> >>> On Wed, May 25, 2022 at 12:21 PM Justin Elser < >>> jus...@or...> wrote: >>> >>>> Hi all, >>>> >>>> I have set up a jbrowse 1.16.11 instance (sorry, not publicly >>>> available) >>>> and added 2 species (well, different cultivars) by doing the following: >>>> bin/prepare-refseqs.pl --out data/nipponbare --fasta >>>> source_files/RAPDB-Reference-Nipponbare_chr_fixed.fasta >>>> bin/flatfile-to-json.pl --out data/nipponbare --trackType >>>> CanvasFeatures >>>> --trackLabel MSU7.0 -gff source_files/MSU7.0.gff3 >>>> bin/generate-names.pl --out data/nipponbare >>>> >>>> bin/prepare-refseqs.pl --out data/arc --fasta >>>> source_files/magic_16/oryza_sativa_arc.fa >>>> bin/flatfile-to-json.pl --out data/arc --trackType CanvasFeatures >>>> --trackLabel Genes -gff source_files/magic_16/oryza_sativa_arc.gff >>>> bin/generate-names.pl --out data/arc >>>> >>>> Then I edited the jbrowse.conf file as explained in >>>> https://jbrowse.org/docs/dataset_selector.html to have: >>>> dataRoot = data/nipponbare >>>> [datasets.nipponbare] >>>> url = ?data=data/nipponbare >>>> name = Nipponbare >>>> [datasets.arc] >>>> url = ?data=data/arc >>>> name = ARC >>>> >>>> And all seems to work fine, except that I don't see the dropdown to >>>> switch datasets, so I can't actually see the ARC dataset. Although I >>>> have by messing around with jbrowse.conf and tracks.conf gotten it to >>>> work, but at this point I've left it out. >>>> >>>> However, the issue I'm having is actually with Apollo. I have the >>>> docker >>>> instance of apollo running and it seems to be fine as configured in a >>>> systemd unit file: >>>> ExecStart=/usr/bin/docker run --rm --name %n \ >>>> -e APOLLO_PATH=apollo \ >>>> -v /data/www/panoryza/jbrowse/data:/data \ >>>> -v /data/www/panoryza/apollo/postgres:/var/lib/postgresql \ >>>> -v >>>> /data/www/panoryza/apollo/apollo_data:/data/temporary/apollo_data \ >>>> -p 8887:8080 \ >>>> gmod/apollo:latest >>>> >>>> Adding Nipponbare as an organism with "Directory" /data/nipponbare >>>> works >>>> fine and I can see the tracks I've added. >>>> >>>> Now, here is where the issue I have starts. If I edit the >>>> data/nipponbare/tracks.conf file as shown in the dataset selector >>>> documentation above: >>>> [general] >>>> dataset_id = nipponbare >>>> >>>> Apollo then shows the following error: >>>> Error message(s): >>>> TypeError: Cannot read properties of undefined (reading 'name'). >>>> >>>> Clearing out the data/nipponbare/tracks.conf file allows apollo to work. >>>> Now, the error only appears in apollo, jbrowse works fine with the >>>> edits. >>>> >>>> I'm a bit confused about how this should be configured to work. I >>>> actually have 14 more "species" to add, but waiting till I at least get >>>> it working with one. The documentation isn't clear to me on the >>>> difference between adding the datasets to jbrowse.conf vs >>>> data/tracks.conf vs data/(species)/tracks.conf. >>>> >>>> Please let me know if you need any more information. >>>> >>>> Thanks, >>>> Justin >>>> >>>> -- >>>> ********************************************************** >>>> * * >>>> * Justin Elser * >>>> * Computational Biology Research Associate * >>>> * Dept. of Botany and Plant Pathology * >>>> * Jaiswal Lab * >>>> * Oregon State University * >>>> * * >>>> * email: jus...@or... * >>>> * * >>>> ********************************************************** >>>> >>>> >>>> >>>> _______________________________________________ >>>> Gmod-ajax mailing list >>>> Gmo...@li... >>>> https://lists.sourceforge.net/lists/listinfo/gmod-ajax >>>> >>> >>> -- >>> ********************************************************** >>> * * >>> * Justin Elser * >>> * Computational Biology Research Associate * >>> * Dept. of Botany and Plant Pathology * >>> * Jaiswal Lab * >>> * Oregon State University * >>> * * >>> * email: jus...@or... * >>> * * >>> ********************************************************** >>> >>> >> -- >> ********************************************************** >> * * >> * Justin Elser * >> * Computational Biology Research Associate * >> * Dept. of Botany and Plant Pathology * >> * Jaiswal Lab * >> * Oregon State University * >> * * >> * email: jus...@or... * >> * * >> ********************************************************** >> >> > -- > ********************************************************** > * * > * Justin Elser * > * Computational Biology Research Associate * > * Dept. of Botany and Plant Pathology * > * Jaiswal Lab * > * Oregon State University * > * * > * email: jus...@or... * > * * > ********************************************************** > > |