Annotations to Cell Fraction-type terms

From GO Wiki
Jump to: navigation, search

UniProt-GOA suggestions

Situation 1. If it was agreed to obsolete the cell fraction terms

a. Remove any annotation that is duplicated by an in vivo CC location (see rules for individual terms below)

b. Remaining annotations would be mapped to an equivalent CC term (following suggestions below) N.B. UniProtKB has only 64 EXP-annotations and 27 NAS/TAS-annotations to cell fraction or children, that are not duplicated by an in vivo CC location

c. Remove ISS and Ensembl Compara IEA predictions.

Situation 2. If the cell fraction terms were not obsoleted

1. Only allow EXP-level annotations to cell fraction and children in future and only where there is no EXP evidence for an in vivo location. Discourage use of such terms in general, encourage use of the 'mapped' terms instead. (Although we are very interested in referring to a separate ontology for cell fraction terms, if this gets implemented).

2. Remove any ISS or any IEAs from Ensembl Compara

3. Remove/disallow manual annotation to cell fraction/children where:

Annotations to CELL FRACTION. Remove if any other CC annotation. Remaining annotations map to GO:cell part

Annotations to SOLUBLE FRACTION. Remove if any other CC annotation. Remaining annotations map to GO:cell part

Annotations to INSOLUBLE FRACTION. Remove if any other CC annotation. Remaining annotations map to GO:cell part

Annotations to MEMBRANE FRACTION. Remove if EXP-level annotations for GO:0016020 membrane or children. Remaining annotations map to GO:membrane

Annotations to PLASMA MEMBRANE ENRICHED FRACTION. Remove if EXP-level annotations for GO:0005886 plasma membrane or children. Remaining annotations map to GO:plasma membrane

Annotations to INTEGRAL TO MEMBRANE OF MEMBRANE FRACTION. Remove if EXP-level annotations for GO:0016021 integral to membrane or children. Remaining annotations map to GO:integral to membrane

Annotations to PERIPHERAL TO MEMBRANE OF MEMBRANE FRACTION. Remove if EXP-level annotations for GO:0019898 extrinsic to membrane or children. Remaining annotations map to GO:extrinsic to membrane

Annotations to SYNAPTOSOME. Remove if EXP-level annotations for GO:0043005 neuron projection or children Remaining annotations map to GO:neuron projection

Annotations to VESICULAR FRACTION. Remove if EXP-level annotations for GO:0031982 vesicle or children. Remaining annotations map to GO:vesicle

Annotations to MICROSOME. Remove if EXP-level annotations for GO:0043231 intracellular membrane-bounded organelle or children. Remaining annotations map to GO:intracellular membrane-bounded organelle

Annotations to ROUGH MICROSOME. Remove if EXP-level annotations for rough endoplasmic reticulum or children. Remaining annotations map to GO:rough ER

Annotations to SMOOTH MICROSOME. Remove if EXP-level annotations for smooth endoplasmic reticulum or children. Remaining annotations map to GO:smooth ER

Annotations to HOST CELL MICROSOME. Remove if EXP-level annotations for host cell intracellular membrane-bounded organelle or children. Remaining annotations map to GO:host cell intracellular membrane-bounded organelle


UniProt controlled vocabularies2GO

There are: 411 IEA UniProt Keyword annotations (169 to Swiss-Prot entries)

1326 IEA UniProt Subcellular Location annotations (1035 to Swiss-Prot entries)

Consider mapping these to the equivalent terms as shown below;

UniProt vocabulary -> GO term

microsome -> GO:intracellular membrane-bounded organelle

microsome membrane -> GO:organelle membrane (currently mapped to GO:microsome)

host microsome -> GO:host cell intracellular membrane-bounded organelle

host microsome membrane -> GO: host cell intracellular membrane-bounded organelle (currently mapped to GO:host cell microsome)

transmembrane -> GO:integral to membrane (currently mapped to GO:membrane fraction)

synaptosome -> GO:neuron projection

MGI

(From Harold)

After discussion our group decided that these fraction terms didn't belong in the Cellular Component branch but the cell fraction experiments could support annotation to specific components.. MGI annotations to these terms came from either our own work or from data loads from either GOA_Mouse or via orthology loads from GOA (human) or RGD (Rat).

MGI have already acted and removed all MGI curator usage of these terms. We used the mappings shown below, keeping the original evidence code (IDA in most cases) and in addition, ECO:0000004 cell fractionation evidence has been entered in our annotation properties/extension table for evidence, We can supply the evidence extension in column 16 if needed. The only remaining usage of these old fraction terms at MGI is now from loads from other groups. We expect these will be modified following group discussion. ___________________

Here are the cell fraction translations we used.

   If  annotation to "Cell fraction"--- no substitute GO term: then remove annotation
   If annotation to "Insoluble fraction" --- no substitute GO term: then remove annotation
   If  annotation to  "Membrane fraction" ---substitute  membrane ; GO:0016020---cell fractionation evidence ; ECO:0000004
   If  annotation to  "Integral to membrane of membrane fraction" —substitute integral to membrane ; GO:0016021---cell fractionation evidence ;
   ECO:0000004
   If  annotation to "Peripheral to membrane of membrane fraction"---substitute extrinsic to membrane ; GO:0019898---cell fractionation evidence ;
   ECO:0000004
   If  annotation to "Plasma membrane enriched fraction" ---substitute plasma membrane ; GO:0005886---cell fractionation evidence ;
   ECO:0000004
   If  annotation to "Synaptosome-" --substitute synapse ; GO:0045202---cell fractionation evidence ; ECO:0000004 (PMID:7903689)
   If  annotation to "Vesicular fraction" ---no substitute GO term, but can often make an annotation based on the individual experiment because they basically show co-purification with markers from known cellular components— for this case add co-fractionation evidence ; ECO:0000102
   If annotation to "microsome" --- no substitute GO term ; then remove annotation
   If  annotation to "Rough microsome"--- no substitute GO term: then remove annotation
   If  annotation to "Smooth microsome"--- no substitute GO term: then remove annotation
   If  annotation to  "Soluble fraction" --- no substitute GO term: then remove annotation

TAIR

(Donghui and Tanya)

(a) We suggest get rid of all these cellular fraction terms. We reviewed the list of TAIR annotations in Rama's spreadsheet, most of the TAIR genes that are annotated to these terms also have EXP annotations to other natural CC terms. A few genes count 'microsome' as the only CC annotation with an EXP code. We could map these annotations to GO:0043231 intracellular membrane-bounded organelle.

(b) Can we also consider keeping (some or all) these celluar fraction terms as synonyms of natural CC terms, e.g. microsome as a synonym for GO:0043231 intracellular membrane-bounded organelle?

PomBase

We don't use them, we would love to get rid of them. Will be confusing for community curation (we intend to "block them, but would prefer if they were removed" Better for annotation consistency.

FlyBase

(Susan)

I am happy for these terms to be obsoleted. I've reviewed our annotations and only a handful genes would lose cellular component information - these can all be mapped to other terms.

Other comments

1. Could use the ECO code ECO:0000004 as evidence for these annotations

[Term] id: ECO:0000004 name: cell fractionation evidence def: "Used when an annotation is made based on separation of subcellular components based on their physical properties, such as density in sucrose density gradients. This evidence is used mostly for annotations to the cellular component ontology." [TAIR:TED]

synonym: "IDA: cell fractionation" RELATED []

is_a: ECO:0000100 ! fractionation evidence

2. Chris's suggestion: Move the cell fraction terms into a separate ontology that would be reference in the 'with' field of a GO annotation to the in vivo cellular component. Have to consider what evidence code would be used as cannot currently have an entry in the 'with' of an IDA annotation.