monarch-initiative / monarch-legacy Goto Github PK

JSONP output from API calls may be needed for AJAX functionality for deployment of widget on non-Monarch sites.

we need a method to fetch orthologs given a gene id
getGeneOrthologs(sourceGeneID,targetSpecies)

I've written a view that integrates the ortholog calls from panther, and some other databases, specifically for human-to-model orthology. It presently includes mouse and fish orthology.

see: http://beta.neuinfo.org/services/v1/federation/data/nlx_152525-14.json?includePrimaryData=true&q=ENSEMBL_ENSG00000160818

Is this the right strategy? Do we want to expand this view to include all ortholog calls between all organisms, or just keep this simple with human-to-other?

Use https://developer.mozilla.org/en-US/Persona

Additional features TBD:

• search history?
• manage sets of entities?

Search box should have autocomplete enabled.

Reference:
https://confluence.crbs.ucsd.edu/display/NIF/OntoQuestMain#OntoQuestMain-1%29Listallontologies

A new api call that, given a genotype id, will return it's composite gene ids. (and symbols?) Or should this be given a genotype id, fetch all the genotype properties? This should presently call on nif-0000-00096-4.

Like: http://beta.neuinfo.org/services/v1/federation/data/nif-0000-00096-4.json?includePrimaryData=true&q=MGI_2166318

Perhaps there should be the ability to query with *, so that the user can dump all genotype-gene mappings, organism specific.

it looks like this might have gotten a start, here: fetchMgiGenotype

autocomplete on SHH takes you to a results page
http://tartini.crbs.ucsd.edu/search/SHH
where the category and Info are "null" and "undefined", though this appears to be a match on a pathway constituent.

Furthermore, why doesn't autocomplete take me to any models with shh in the label, if it is currently string based (rather than the semantic entity "shh gene".

Clicking on the "model" column under models for a disease leads to this link:

https://secret-harbor-1370.herokuapp.com/model/FBInternalGT:254236,

which leads to

TypeError: Cannot find function fetchModelInfo in object [object Object].

lib/monarch/web/webapp.js, line 400

Script Stack

at lib/monarch/web/webapp.js:400 (anonymous)
at stick/lib/middleware/route.js:215 (route)
at stick/lib/middleware/params.js:102 (anonymous)
at stick/lib/middleware/static.js:62 (static)
at stick/lib/stick.js:37 (app)
at ringo/jsgi/connector.js:42 (handleRequest)

This tab should have two sub-selections (for now).

The first selection is the "Monarch tour" from the Contact tab, which should be relabeled "Monarch Info" or "Monarch Information"

The second is "Releases" and we should move the release info from the "Monarch tour" section, but this needs to be updated.

Each release should have a unique identifier. About pages should have appropriate ID #s for ontologies releases.

Need API call that can fetch all models annotated with a given MP term. (This, or something like it, may already be implemented.)

I think first iteration we will probably have to hardcode this to a set of views. At first nif_0000_21427_10 for mouse and nlx_151835-1 for human disease, but eventually we will need this to be a bit more loosey and do a generic federation query and filter the results on "genotype" concepts.

Wrap:
http://beta.neuinfo.org/services/v1/federation/data/nlx_152525-9.tsv?q=*&offset=0

The REST URL should end with .gff3

We should consider the idea of a search box where the user can randomly select a phenotype, disease, or gene and see what they get, to explore the system.

(just noting the idea- low priority)

This should have a parameter that specifies the number of models to return by rank (default to 10). Because we have several scoring mechanisms, this will default to just one of them (maybe BMAasymIC), but eventually should be configurable.

Initially this is clearly going to fetch just mouse models. But in the future we'll probably need to parametrize the target species.

This will need to draw on nlx_152525-10 to get the mouse models by rank.

This should return a list of (model id, label, score, metric). (The individual attribute comparisons is supplied by another call...although it could technically be all rolled into one if we wanted it to.)

confirm use of monarch theme colors and fonts as designed by N. Skillen.

Ideally this would feed right into this git tracker.
One consideration is that the user may or may not want to provide an email, or otherwise be logged in. Further, most will not be git users so would be anonymous tickets even though we'd want to ideally have their contact info.

On most pages we have the ability to download JSON or n-triples. This should be moved to a more regular consistent place where it isn't inline with the other disease/model/phenotype biology-related content.
for example http://tartini.crbs.ucsd.edu/disease/OMIM_168600

Either have the user's browser open up an email in their client to [email protected], or have a separate page with this email address as well as the PIs.

I prefer not to have a submit message box, as I personally find those rather annoying since they are not in your email history.

Current page has asserted models:
http://secret-harbor-1370.herokuapp.com/disease/OMIM_143100#model

And below that a "similar diseases" section that needs improved..

There should be an additional section showing the inferred models.

@nlwashington - what is the nlx ID for this view?

All (at least human) variants should be displayed in HGVS notation.

I think this should be the default for human mutations, and perhaps a configurable option (by the user) for other species.

In order to properly rank and markup individual attributes (ontology classes) in chuck's grid view (and probably when displaying under the phenotype section), we need the IC scores for the attributes in the annotations.

This data has now been added to the owlsim server, but we need a wrapper function to fetch them. Not sure if this should be a new api call (get score for attribute), or if it should just be built in to any function so that any time an ontology class is fetched, the IC of the class is automatically spit out. This has the unfortunate side effect (at the moment) of possibly requiring two underlying API calls - one to NIF services (either DISCO or OntoQuest), and then a query to our owlsim server for the scores. Maybe that's no big deal, but will certainly add latency time.

One possible query function is simply getICScoresForAttributes(A1...n), where you feed it either a single or a set of attributes and you get back that same set plus scores in JSON.

I think the get LCS scores function ought to return the IC values of the individual classes, as well as the LCS scores.

Allow logging in.

See:
http://ringojs.org/tutorial/sessions.md

add NIH funding acknowledgment to about page and footer.

Should wrap this:
http://code.google.com/p/owltools/issues/detail?id=79

See below for output; when we iterate through all disorders closures are queried making this v slow.

Best short term approach is to turn off closure in this case.

Longer term we shouldn't need to iterate if we just need labels

FETCHING: http://beta.neuinfo.org/services/v1/federation/data/nif-0000-00096-6.json data=includePrimaryData=true&count=1&q=*&project=effective_genotype_name&project=effective_genotype_label&filter=effective_genotype_id:MGI_2175756&
URL: http://beta.neuinfo.org/services/v1/federation/data/nif-0000-00096-6.json?includePrimaryData=true&count=1&q=*&project=effective_genotype_name&project=effective_genotype_label&filter=effective_genotype_id:MGI_2175756&
RESULT: [object Object]
STATUS: 200
Fetched: {"id":"MGI_2175756","label":"Cdk5r1<tm1Lht>/Cdk5r1<tm1Lht> [involves: 129S4/SvJae * C57BL/6]","source":"nif-0000-00096-6","resource":"nif-0000-00096-6"}
FETCHING: http://alpha.neuinfo.org/services/v1/federation/data/nlx_151835-1.json data=includePrimaryData=true&count=1&q=*&includeSubclasses=true&project=disorder_name&project=disorder_label&filter=disorder_id:OMIM_604213&
URL: http://alpha.neuinfo.org/services/v1/federation/data/nlx_151835-1.json?includePrimaryData=true&count=1&q=*&includeSubclasses=true&project=disorder_name&project=disorder_label&filter=disorder_id:OMIM_604213&
RESULT: [object Object]
STATUS: 200
Fetched: {"id":"OMIM_604213","label":"Chudley-Mccullough Syndrome","source":"nlx_151835-1","resource":"nlx_151835-1"}
CACHE undefined

some pages return an error similar to the following:

Failed to load resource: the server responded with a status of 404 (Not Found) http://localhost:8080/css/images/ui-bg_flat_75_ffffff_40x100.png

I think this has something to do with the way that jquery-ui is being used. It should probably be installed from scratch as per instructions.

Use nif-0000-21427-10

Please remove Carlo, he's no longer with us, unfortunately.

For those of us labeled as PIs, perhaps change our bios to read "X is currently a PI of the Monarch Initiative...." that way we can all have equal importance.

For OHSU, please edit my bio to read:
Dr. Haendel has a BA from Reed College in Chemistry and a Ph.D. in Neuroscience from the University of Wisconsin, Madison. She is currently a PI of the Monarch Initiative, with the aim of providing integrated access to model systems phenotype-genotype data for the purposes of disease hypothesis exploration. She also participates in development of the eagle-i and CTSAconnect ontologies, designed to collect and disseminate information about biomedical resources and enable research profiling. Her research interests are in using ontologies to promote synthetic science though connections within biomedical data, to utilize information science during the course of research and its publication, and to enable scientific reproducibility.

Please add Bryan Laraway and Kent Shefchek to OHSU's section.

Please add Jeremy Espino to the UPitt section.

Please add Sarah Kim (yourself!) to the LBNL section.

We'll broadcast requests for any changes to the other bios.
Should we list the HMC team? I think that would be nice.
Also, should we have a contact us button on this page? that gets sent to [email protected]?

For example, http://tartini.crbs.ucsd.edu/disease
choose Smith-Lemli-Opitz syndrome
and go to this page, with no label:
http://tartini.crbs.ucsd.edu/disease/OMIM_270400
the ID shows instead.

The Download buttons take up a lot of real estate. We need to change that.
For example:
http://tartini.crbs.ucsd.edu/phenotype/MP_0005449,
on the genotype tab, it means the data doesn't show up but half-way down the browser.
Here's some ideas for changes:

1. make the download a small tiny link/buttons, next to the table header. I think a small tiny icon for each format would be sufficient (no need for the text that yells at me to download).
2. allow menu option on the tab to download the data in that tab, in various formats.
3. make it an option on the left-bar

List of synonyms should be presented in a more structured way. Add synonym provenance (requires OQ to return this info).

The IDs should be split by database and hyperlinked.

Can you change the regular text (non-header) font to be something easier to read, like helvetica? (That's what I did for my posters.) It is fairly compatible with the Kelso font style-wize, and way easier to read when small.

http://f1000research.com/articles/2-30/v2

We want to be able to consolidate and/or make visually apparent some kind of higher-level group that an annotation term belongs too. In the first iteration, we'll hardcode the grouping to be the 1st level down of HP/Uberpheno.

There needs to be an API call that returns the group/parent or whatever of a term, which itself leverages OQ.

E.g.

/class/:id.json

Options:

/class/:id.json?level=2

Straightforward wrap of:
http://secret-harbor-1370.herokuapp.com/docs/files/api-js.html#bbop.monarch.Engine.fetchClassInfo

URLs are passed along raw without sanitization to clear out potential security risks. Sanitization calls to prevent XSS would be helpful

given a (list of) gene id, i want to see the distribution of orthologs across species, the and the fraction of orthologs that have phenotypes in each species.

a generalization of this should be presented in a graphical way on the front page of monarch, i think.

Stick is non-uniformly escaping html, and causing all sorts of problems with genotypes and alleles during display. This will need to get tracked down and fixed somehow.

things that end up escaped are omim variation ids, like "ABC < OMIM:12345.0001 > ", but not ABC < rs1234566 > .

similarly, almost any mouse model shows up wrong...whatever is in between the < > doesn't want to show up.

Add a robots.txt with a stanza to allow indexing, with a delay. cc @kltm @harryhoch

When you are on http://tartini.crbs.ucsd.edu/phenotype/
you see "Phenotype spotlight" "Large floppy ears". The "Find out more" button goes to an empty phenotype page.

Also, instead, consider making "large floppy ears" clickable as suggested for homepage #50
or at least we should be consistent with how we display the find out more links/buttons across the different pages

Could I get a list of the sources I need to make images for as well as a way to replace the strings with the images (API call?).

See:
https://support.crbs.ucsd.edu/browse/LAMHDI-140

I have built several views in DISCO that we can now query via services, in order to build the JSON necessary to paint our pathway diagrams.

The nif views from KEGG (in beta) are:

disease-to-pathway mappings (also contains omim mappings for disease identifiers)
http://beta.neuinfo.org/mynif/search.php?q=*&t=indexable&nif=nlx_31015-2
(canonical) pathway to KO (kegg ortholog) class
http://beta.neuinfo.org/mynif/search.php?q=*&t=indexable&nif=nlx_31015-3
KO to species gene map (orthology map).
http://beta.neuinfo.org/mynif/search.php?q=*&t=indexable&nif=nlx_31015-4

If I am missing something, please let me know!

What this also means is that we also have enough information to have any kind of pathway page. Do we dare make those too?

Jeremy, what kind of API calls do you need? Please define them here. My guess is:

getGenesForPathway(pathway_id,species);
getPathwaysForGene(gene_id);
getPathwaysForDisease(disease_id);

the two above could generically call
getPathwaysForID(id)
but then we'd need an id-to-type resolver

This would be a function that will wrap whatever data we have...presently will be hardcoded to KEGG, but can wrap wikipathways at somepoint too.

http://owlsim.crbs.ucsd.edu//compareAttributeSets data=a=HP:0000006&a=HP:0000496&a=HP:0000716&a=HP:0000726&a=HP:0000751&a=HP:0001347&a=HP:0002063&a=HP:0002067&a=HP:0002072&a=HP:0002171&a=HP:0002529&a=HP:0000708&a=HP:0001276&a=HP:0001608&a=HP:0002120&a=HP:0002353&a=HP:0002376&a=HP:0100022&b=MP:0002169

An empty array is returned.

On the home page, we (nicely) have an overview of four types of entities in the system. Consider removing the "more details" link, which goes to more details of the main entity (in the attached example, "Parkinson's disease" for the disease box, and instead make the "Parkinson's disease" itself clickable and remove the "more details".

Also consider changing title for each box to just the title of that type "Diseases" instead of "Diseases:Parkinson's disease". Instead, move the "Highlighted Diseases" title to below that, making Parkinsons the first expanded entry, and the others following that. In this way it will be apparent that these are simply a few of the highlighted diseases we have (and that Parkinson's is one of them).

E.g.

The current "compare" api call only includes the individual annotations, and their LCS. We should add the comparison/similarity score for the entirety of the set as well. We should add all relevant scores.

This will draw on the precomputed similarity tables loaded into DISCO in nlx_152525-10.

I finished the skeleton for the tour page, so I am requesting descriptions and images/infographics. These are the sections which probably need information/images:

• Purpose: I just took the text from the Drupal site.
• Beta Release: There is nothing here at the moment except a sentence about contacting [email protected] for future beta testers.
• Future Release: I thought having future models or something here would be cool. A description would be nice, too.

P.S. You can send high quality images to [email protected] or place them in the images folder. Thanks!

doid/merged.owl has equivalence axioms, e.g.

• DOID_123 EquivTo OMIM_987

Annotation assertions are typically only associated to one of these.

When resolving to a page, DOID_123, should redirect to OMIM_987

The left-hand column is non-static in size, and moves around when i make my browser window large. This should just be made static-size on the left-hand side-justified.

It seems that only a fraction of the results are showing up on the disease pages (probably true elsewhere).

For Huntington's disease, none of the mouse models are showing up:
http://secret-harbor-1370.herokuapp.com/disease/OMIM_143100

even though they are clearly there:
http://beta.neuinfo.org/mynif/search.php?q=OMIM_143100&t=indexable&nif=nlx_152525-2&b=0&r=20&ff=Organism:mus%20musculus*Disease%20Model%20Type:asserted