Modify Reactions column in molecules grid

[ridderl]

Example:

• Reactions
  • Reactants
  • Rule1 (nr of molecules/nr of molecules with peak)
  • Products

Clicking will filter metabolites on molecules which are products of rule1.

[sverhoeven]

I need some clarifications on the numbers and what clicking should do. The example is the column content shown for a molecule row. Let's say the molecule has metid 3 and Rule1 is 'esterase'.

• What does nr of molecules mean?

  SELECT count() FROM reactions WHERE name='esterase' or SELECT count() FROM reactions WHERE name='esterase' AND reactant=3

• What does nr of molecules with peak mean?

  Nr of molecules which have been fragmented or Nr of molecules which have been assigned to a peak

• Clicking where will do what? There are 4 places where you can click

  1. 'Reactants' or 'Products' text
  2. 'Rule1'
  3. 'nr of molecules'
  4. 'nr of molecules with peak'

  In the issue description says clicking on 'Rule1' (option 2). Is this correct? Should clicking on option 1 also perform a filtering?

[ridderl]

• For reactants: SELECT count() FROM reactions WHERE name='esterase' AND product=3, for products: SELECT count() FROM reactions WHERE name='esterase' AND reactant=3
• The ones with metabolites.nhits > 0 (where metabolites.metid = reactant or product, resp.)
• Options 1 and 2 are both useful

[sverhoeven]

In issue208 branch there is now functionality for filtering molecules on reactions.

In the column header filter you can filter on:

• reactant, filters molecules on products of reactions with selected reactant (==metid)
• product, filters molecules on reactants of reactions with selected product (==metid)
• name, filters molecules on reactants of reaction with selected product and reaction name or filters molecules on products of reaction with selected reactant and reaction name

[sverhoeven]

The Reaction column is now rendered from json.

To use the metabolites.reactionsequence column in results.db must be refilled. This can be done with web/magmaweb/models.py:fill_molecules_reactionsequence() with:

    from magmaweb.job import JobFactory
    factory = JobFactory('data/jobs')
    session = factory._makeJobSession('58f05077-aad8-4fc9-a497-310495ab8b62')
    from magmaweb.models import fill_molecules_reactionsequence
    fill_molecules_reactionsequence(session)

[sverhoeven]

To update job result databases use:

cd web
pip install alembic
for x in `ls data/jobs` ; do echo $x;alembic -x jobid=$x upgrade head; done

[sverhoeven]

The reaction filter can give zero results because of other filters like nhits>1 and scanid.

The Failed to load fragments from server error is back. When a scan has been selected and then selecting a reaction (which has molecules with a scan) will give the error.

[sverhoeven]

Errors occurs only when reaction has 1 molecule as target. Before click a scan and molecule are selected, after click same scan is still selected, but another molecule is selected which did not form fragments on selected scan.

Solution: Scan markers are filled check if previously selected scan is absent then unselect scan