Hello. After using MitoMiner for a while, we've come up with a few questions about customising some of the viewable fields, and a few possible bugs. I was also wondering if the UniProt parser sequence checker issue was fixed yet.
dev mailing list
Sorry for the delay in replying, I was in California last week.
1. In InterMine 1.0 there will be another parameter you can set in your project
XML to allow for duplicate sequences.
2. No, it's not possible to hide constraint options but you shouldn't be getting
errors. Can you send me the failing query XML and the error message you get?
3. Yes, that's a known problem, others have brought this up as well. It's in
the queue of things to fix but it's not been done for 1.0.
4. No, this isn't something you can currently configure. You can change this
manually by changing line 129 of headMenu.jsp:
These two items comprise the submenu for the list page:
Switch the order and the "view lists" link will be the default:
5. No, it is not currently possible to hide fields from the "add columns" form.
However it makes sense to me to hide columns that have no values or at least
have the option to do so or indicate that it's empty in some way. Let me make a
ticket and put this in the queue. This won't be done for 1.0 though, I think.
I hope that answers your questions.
You were also going to change the mitominer model to match more closely with the
InterMine core model. Were you able to do that successfully? You'll need that
done before you upgrade to 1.0 to take advantage of the updated UniProt parser.
I was also going to help you write a post-process (?) or some new code to move
the mitominer specific code outside of the UniProt parser. This was to make
debugging easier (or even possible!). I think this was to do with the
isMitochondrial boolean in your Protein class.
Let me know how you're progressing. I'm guessing it's going to be a big job!
On 01/05/12 16:49, James Blackshaw wrote:
> Hello. After using MitoMiner for a while, we've come up with a few questions
> about customising some of the viewable fields, and a few possible bugs. I was
> also wondering if the UniProt parser sequence checker issue was fixed yet.
> * If we use a negative constraint on certain flags in MitoMiner (i.e.
> MitoEvidence_mitoCarta has no value), we get an error. If we
> constrain to false we get no results as we just don't populate the
> field if the evidence isn't there. Is it possible to hide the
> constraint options that make no sense to use?
> * Summarising a column in the results table, when that column contains
> only numbers, seems to trigger some statistics (Min, Max, Sample
> Mean, Standard Deviation) rather than an actually useful summary,
> such as cumulative frequency.
> o Example Query: <query name="" model="genomic"
> MassSpectrometry.publication.title" longDescription="Show
> proteins that have mitochondrial mass spec evidence, for a
> specific species"
> sortOrder="MassSpectrometry.protein.UniProt_accession asc">
> <constraint path="MassSpectrometry.protein.organism.long_name"
> op="=" value="Mus musculus"/> </query>
> * Can we set the Lists page to display "View" instead of "upload" as
> the default sub-section, so that users get to see the example lists
> we already make available?
> * Can we hide fields from the "add column to results" field as well as
> the query builder? For example, "name and symbol" don't have values
> so we hide them. They are however selectable as columns to add to a
> dev mailing list
> [hidden email]
dev mailing list
|Powered by Nabble||Edit this page|