Advance Search First Test

[ypriverol]

Hi @baimingze firt of all thanks a lot for this work. The tool work perfectly fine and it will help our users a lot to find and search for data. Some minor comments before it integration in the release branch:

• In my google chrome it doesn't work. I tested in Firefox to see the final results. (version 55.0.883).
• Keep the search query value. Imagine I search in the main page for a big query and i would like to edited in the result page (see Figure 1), I think if the user select advance search in the result it should be easy for us to populate the advance search box in order to edit the fields.
• The box as some pixels out not aligned with the parent box (see figure 2)
• I think we should remove the grey color of the advance search and keep only white. (figure 3)
• The most important feature is how to query for reference fields for example uniprot ids,
  ensembl or chebi ids, this is what most of the users are looking for a way to search for all
  datasests where two . n particular proteins are found. Is possible to implement this feature.

Figure 1

Figure 2

Figure 3

[baimingze]

hi @ypriverol , could you give more details about chrome not working isssue? Like see what the console told by pressing F12.

[ypriverol]

http://www.ebi.ac.uk/ebisearch/ws/rest/{domain}/xref http://www.ebi.ac.uk/Tools/webservices/services/eb-eye_rest#cross-reference_search

[baimingze]

I guess we already can search by uniprot ID?

[ypriverol]

http://www.omicsdi.org/search?q=UNIPROT:(P04637)%20AND%20UNIPROT:(P01106)%20AND%20TAXONOMY:%229606%22

We should be able to filter by the cross-reference database and the Key. For example UNIPROT and then the accession. like the query

[mindcrusher11]

Instead of condition we can have fields like in ncbi

[ypriverol]

Some other tips we should take into account:

• In the Advance box you should expect to have instructions about what needs to be done and about the options filters. http://www.getelastic.com/wp-content/uploads/as-where.jpg we should avoid hijacking searchers to a new page or pop up window; to get detailed instructions by using Javascript hover effects or “plus-box” expand/collapse.

[baimingze]

feedbacks from Henning:

• typo: advanced
• get the query already in search into the advanced search
• some example for the advanced search
• open on condition for the user
• another button in the condition trees
• well document

[ypriverol]

@Mingze:
If I understand properly, we can do some priorities and continue release in steps, Priority:

• Typos
• Instead of conditions, is better to renamed to fields as suggested by @mindcrusher11
• Been able to use cross-references because 90% of the users want to see the proteins and
  genes.
• css problems
• add tooltips to all buttons with small explanation about what is each thing.

We can than continue with the more complex things:

• been able to represent the simple query into the advance box including between pages.
• documentation.

[baimingze]

@ypriverol Agree

[baimingze]

@ypriverol I can get the Cross Reference Fields from here: http://www.ebi.ac.uk/ebisearch/ws/rest/omics ,
which is inspired from your info, but not from "http://www.ebi.ac.uk/ebisearch/ws/rest/omics/xref", which provides nothing.

I still can not find a way to list all the Uniprot IDs for user to select, do you have some ideas about this? And I also wondering, the list should be a huge list with only ids, it may not a good idea to list them. How do you think?

[ypriverol]

First if I don't understand wrong you can get the cross-reference from http://www.ebi.ac.uk/ebisearch/ws/rest/omics then get it from there?

The second question, when you type P04637 in the general search is found by the auto-complete. Then (inprinciple) it must be possible to restrict the autocomplete to only the Uniprot field and get the list of proteins with auto complete?.

[baimingze]

                var xref_fields = [
                    {name: 'UNIPROT', label:'UNIPROT'},
                    {name: 'PUBMED', label:'PUBMED'},
                    {name: 'ENSEMBL', label:'ENSEMBL'},
                    {name: 'WORMGENE', label:'WORMGENE'},
                    {name: 'PUBCHEM', label:'PUBCHEM'},
                    {name: 'CHEBI', label:'CHEBI'},
                    {name: 'HMDB', label:'HMDB'},
                    {name: 'KEGG', label:'KEGG'},
                    {name: 'NCBI', label:'NCBI'},
                    {name: 'PASS', label:'PASS'},
                    {name: 'SGD', label:'SGD'}
                ]

I collected them from http://www.ebi.ac.uk/ebisearch/ws/rest/omics , with the

<option name="type">XREF</option>

from

<fieldInfo id="PUBMED" name="PUBMED">
<options>
<option name="searchable">true</option>
<option name="retrievable">true</option>
<option name="sortable">false</option>
<option name="facet">false</option>
<option name="referenced domain">medline</option>
<option name="referenced field">id</option>
<option name="type">XREF</option>
<option name="topterms">false</option>
</options>
</fieldInfo>