geneontology / go-ontology

Source ontology files for the Gene Ontology
http://geneontology.org/page/download-ontology
Creative Commons Attribution 4.0 International
220 stars 40 forks source link

EC/GO mapping #1279

Closed gocentral closed 9 years ago

gocentral commented 20 years ago

Some EC numbers correspond to more than one GO term. I have examined the cases where this occurs and have the following recommendations:

EC:1.1.1.1 - alcohol dehydrogenase activity %alcohol dehydrogenase activity, iron-dependent %alcohol dehydrogenase activity, metal ion-independent %alcohol dehydrogenase activity, zinc-dependent

Do the children really represent a different activity to the parent term? Surely this refers to protein structure or to reaction conditions. For the time being, will remove the dbxrefs from the children (since they're more specific) but I think these terms need reassessing.

EC:1.15.1.1 - superoxide dismutase activity %copper, zinc superoxide dismutase activity %iron superoxide dismutase activity %manganese superoxide dismutase activity %nickel superoxide dismutase activity The children are differentiated from the parent on the basis of certain structural features / reaction conditions. See the comments for EC:1.1.1.1 above.

EC:1.16.3.1 - ferroxidase activity %multicopper ferroxidase iron transport mediator activity The latter is a nasty, nasty gene product which should be obsoleted. Transfer annotations to iron transport and ferroxidase activity.

EC:1.18.6.1 - nitrogenase activity %iron-iron nitrogenase activity %molybdenum-iron nitrogenase activity %vanadium-iron nitrogenase activity The children are not functionally different to the parent term. Since corresponding component terms exist for the children, I suggest obsoleting these function terms and annotating to the parent and the corresponding component term.

EC:1.18.99.1 - TRANSFERRED! Needs to be updated.

EC:1.3.5.1 - %quinol:fumarate oxidoreductase activity %succinate dehydrogenase (ubiquinone) activity This is the same reaction but going in opposite directions, so both terms are equally deserving of the EC ID. Reaction direction shown by EC is the latter.

EC:2.4.2.22 - %xanthine phosphoribosyltransferase activity %xanthine-guanine phosphoribosyltransferase activity These look to be identical and should probably be merged.

EC:2.7.1.123 - kinase nastiness. %calcium/calmodulin-dependent protein kinase activity %calmodulin-dependent protein kinase I activity The first term is the EC recommended name -> prob should remove the dbxref from the other term.

EC:2.7.1.37 - this is protein kinase, but we have cAMP-dependent protein kinase activity protein kinase CK2 activity protein serine/threonine kinase activity ...mapped to it. These are specific instances of the activity so the dbxref should prob be removed and applied to 'protein kinase activity' instead. This whole section needs to be looked at in greater detail.

EC:2.7.7.7 - This is the DNA-directed DNA polymerase term and all its children - alpha, beta, etc., DNA polymerases. These children should all be examined in great detail at some point because they all sound suspiciously like gene products. For the time being, am removing the dbxrefs from the child terms.

EC:3.1.26.4 -ribonuclease H activity %ribonuclease H1 activity The child actually seems less specific than the parent from the def! All the nucleases really need to be overhauled because they are all very gene product-y.

EC:3.1.3.16 - protein kinases. Need to be looked at further because they are all suspiciously gene product-y.

EC:3.1.3.48 - protein tyrosine phosphatases Have removed the dbxref from the specific child but all the children could do with being examined further.

EC:3.6.1.47 - small monomeric GTPase activity %various specific (ARF/RAB/RAN, etc.) smGTPases These really need some serious looking at because the children do not differ functionally from the parent. For the time being, remove the dbxrefs but these terms could do with being reassessed <cough obsoleted cough>.

EC:3.6.1.48 - protein-synthesizing GTPase activity %protein-synthesizing GTPase activity, elongation %protein-synthesizing GTPase activity, initiation %protein-synthesizing GTPase activity, termination These are process-function crossovers - even the parent term is rather dubious (and its def is very gene product-y). Suggest obsoleting the child terms and directing the annotations to the parent plus a process term.

EC:3.6.3.21 - polar-amino-acid-transporting ATPase activity %arginine porter activity -- OK; removed dbxref %cystine/diaminopimelate porter activity -- split this term %glutamate/aspartate porter activity -- split %glutamine porter activity -- OK; removed dbxref %histidine/arginine/lysine/ornithine porter activity -- split

EC:3.6.3.22 - nonpolar-amino acid-transporting ATPase activity %leucine/isoleucine/valine porter activity

EC:3.6.3.32 - quaternary-amine-transporting ATPase activity %glycine betaine/proline porter activity The child terms are combined transporter terms representing a specific TC transporter. Recommend obsoleting the child term as it covers more than one function and annotating to transporter terms for each substrate.

EC:4.6.1.1 - adenylate cyclase activity %calcium/calmodulin-responsive adenylate cyclase activity Represents reaction conditions / gene product properties.

EC:5.2.1.8 - peptidyl-prolyl cis-trans isomerase activity %cyclophilin-type peptidyl-prolyl cis-trans isomerase activity The child term is differentiated from the parent on the basis of gene product characteristics. Recommend obsoleting the child; annotation should be to the parent 'peptidyl-prolyl cis-trans isomerase activity'.

EC:5.4.2.1 - phosphoglycerate mutase activity %2,3-bisphosphoglycerate-dependent phosphoglycerate mutase activity %2,3-bisphosphoglycerate-independent phosphoglycerate mutase activity The children (which have 0 annotations) represent specific reaction conditions. Recommend merging these terms into the parent.

EC:6.3.1.5 - NAD synthase activity %NAD synthase (AMP-forming) activity %NAD synthase (glutamine-hydrolysing) activity (EC:6.3.5.1) The parent seems to have been added as a grouping term for the latter and 'NAD synthase (glutamine-hydrolysing)'. Since the defs of NAD synthase and NAD synthase (AMP-forming) are identical, I would recommend merging these terms and having NAD synthase (glutamine-hydrolysing) as a sibling.

EC:4.2.99.18 - will examine in more detail / consult with an expert.

Reported by: jl242

Original Ticket: "geneontology/ontology-requests/1282":https://sourceforge.net/p/geneontology/ontology-requests/1282

gocentral commented 20 years ago

Logged In: YES user_id=835115

Amelia, The suggested changes look fine to me, however being a novice as I am, I would encourage you to exercise your best judgement where major changes to the graphs are required.

To be sure all the offending records are on record....Including your suggested changes above, I think the following EC#s will still need attention: EC 1.1.3.15 1.18.1.2 1.21.3.4 2.4.1.1 2.7.7.49 3.1.1.1 3.1.1.32 3.1.1.4 3.1.2.14 3.1.3.5 3.1.3.66 3.1.3.67 3.2.1.14 3.4.16.5 3.4.22.36 3.6.1.3 3.6.1.6 3.6.3.22 4.2.99.18 4.6.1.2 6.3.1.5

Be sure to let me know if there is anything I can do to help! -Doug

Also, I have reconsidered two of the mappings in the Excel file I sent you. I think EC:1.1.3.15 could map to GO:0003973 and EC:1.18.1.2 could map to GO:0004324, rather than the GOs I had previously suggested.

Is there a way to maintain these customizations to the ec2go file from one production of ec2go to the next?

-Doug

Original comment by: doughowe

gocentral commented 20 years ago

Logged In: YES user_id=451873

Hi Doug,

I've already dealt with most of the EC numbers in your list; they were simpler cases where one of the reactions was more specific than that given in EC, or where one of the terms was obsolete. If everyone agrees to the changes I have proposed, I will implement them next week; the ec2go mapping can be regenerated then and we will hopefully have eliminated the problem of several GOids mapping to a single EC id.

Thanks again for your help, Amelia.

Original comment by: jl242

gocentral commented 20 years ago

Logged In: YES user_id=473796

Have corrected most of the EC numbers mentioned above; many were specific children from which I removed the dbxref; the others were: EC:1.16.3.1 - multicopper ferroxidase iron transport mediator activity obsoleted EC:1.18.6.1 - children obsoleted EC:1.18.99.1 - corrected the EC number and removed the dbxref from the specific child terms EC:2.4.2.22 - merged the terms since they were identical EC:3.6.1.48 - children obsoleted EC:3.6.3.21, 22, 32 - multifunctional children obsoleted EC:5.2.1.8 - child obsoleted EC:6.3.1.5 - terms merged EC:4.2.99.18 - have left the EC ref on 'GO:DNA-(apurinic or apyrimidinic site) lyase activity ; GO:0003906' - the other term was more specific

I haven't obsoleted or merged any of the other terms as I am not convinced anyone looked at this SourceForge item! I will bring them up in another item.

The only undealt with item is EC:1.3.5.1, which represents the same reaction but in opposite directions. In this case I am going to leave it as it is, possibly pending further discussion.

Original comment by: girlwithglasses

gocentral commented 20 years ago

Logged In: YES user_id=473796

Added SF:858130 which deals with EC:1.3.5.1.

Will bring up those other items at another time.

Original comment by: girlwithglasses

gocentral commented 20 years ago

Original comment by: girlwithglasses