Rhea annotations #241
Rhea annotations #241
Conversation
and submit a supporting table to justify the newly added IDs. See issue #224 for more information about the mapping.
Rhea master ID correspond to the same reaction but having an undirected equation direction. For some line the rxnRheaID == rxnRheaMasterID
This file has been used to map Rhea reaction equations to Human-GEM equations
- there was missing matching for some transport reactions - some IDs were not written in reactions.tsv when there was multiple RheaIDs matching the same reaction
model/reactions.tsv
Outdated
| @@ -6024,11 +6024,11 @@ rxns rxnKEGGID rxnBiGGID rxnEHMNID rxnHepatoNET1ID rxnREACTOMEID rxnRecon3DID rx | |||
| "HMR_5984" "" "" "" "r2069" "" "r2069" "MNXR105978" "HMR_5984" "RCR40526" "2.A.3.8.15" 0 "MA-R05984" "" "" | |||
| "HMR_5985" "" "" "" "r2070" "" "r2070" "MNXR105979" "HMR_5985" "RCR40160" "2.A.3.8.15" 0 "MA-R05985" "" "" | |||
| "HMR_5986" "" "" "" "r2071" "" "r2071" "MNXR105980" "HMR_5986" "RCR41449" "2.A.3.8.15" 0 "MA-R05986" "" "" | |||
| "HMR_5987" "" "FE2t" "" "r2072" "" "FE2DMT1" "MNXR99505" "HMR_5987" "RCR40527" "2.A.55.2.3" 0 "MA-R05987" "" "" | |||
| "HMR_5987" "" "FE2t" "" "r2072" "" "FE2DMT1" "MNXR99505" "HMR_5987" "RCR40527" "2.A.55.2.3" 0 "MA-R05987" "RHEA:29441;RHEA:29440;RHEA:29581;RHEA:29580" "RHEA:29439;RHEA:29579" | |||
There was a problem hiding this comment.
Here, the Rhea master IDs can be difficult to associate programmatically to the Rhea IDs (2 master IDs versus to 4 Rhea IDs). Do you think the content should be changed, for instance by always having the same number of IDs in the two columns (this implies to duplicate Rhea master IDs)?
There was a problem hiding this comment.
Another solution would be to manually curate the association now, some Rhea IDs can be probably removed and it might fix the problem.
There was a problem hiding this comment.
Among the four Rhea ids mapped to HMR_5987, only RHEA:29581 appears to be the correct association while the others can be ruled out according the reaction direction or metabolite compartment.
model/reactions.tsv
Outdated
| @@ -8426,7 +8426,7 @@ rxns rxnKEGGID rxnBiGGID rxnEHMNID rxnHepatoNET1ID rxnREACTOMEID rxnRecon3DID rx | |||
| "FAS80COA_L" "" "FAS80COA_L" "" "" "" "FAS80COA_L" "MNXR99415" "" "" "" 0 "MA-R01036" "" "" | |||
| "FBA5" "" "FBA5" "" "" "" "FBA5" "MNXR99463" "" "" "" 0 "MA-R01038" "" "" | |||
| "FCLTm" "" "FCLTm" "" "" "" "FCLTm" "MNXR99471" "" "" "" 0 "MA-R01044" "RHEA:22584" "RHEA:22584" | |||
| "FE2tm" "" "FE2tm" "" "" "" "FE2tm" "MNXR99501" "" "" "" 0 "MA-R01046" "" "" | |||
| "FE2tm" "" "FE2tm" "" "" "" "FE2tm" "MNXR99501" "" "" "" 0 "MA-R01046" "RHEA:29441;RHEA:29440;RHEA:29581;RHEA:29580" "RHEA:29439;RHEA:29579" | |||
There was a problem hiding this comment.
Same problem for that reaction.
|
The most recent commit introduced some new associations that include one-to-multiple links between Human-GEM and Rhea. This is completely fine if the links can be verified programmatically or manually. Otherwise, they probably should be left out. From the point of view of reviewer, it's alright to have less associations as long as they are accurate. |
|
One-to-many links increase the difficulty of using the annotations as they need to be treated differently in code. If possible, sticking to a single id simplifies the work. |
remove association when the extracellular metabolite "X[s]" was not assigned to the metabolite "(out)" in the Rhea equation
but keep these associations in the rhea_reactions_association.tsv for later curation
|
@pecholleyc it's pleasant to go through these Rhea associations, especially those from the approach via equation mapping that nicely link to the ones with proper reaction direction. While the way through UniProt link to Rhea ids in undefined direction. Could the latter be revised with proper direction by using the equation mapping method? Anyway, this PR is good enough as it is. |
The Rhea IDs obtain using UniProt were only Rhea master IDs (undirected reactions). Here the corresponding IDs for bidirectional reactions were automatically updated, and the corresponding IDs for directed reactions were manually curated.
|
@pecholleyc were these Rhea IDs in 8d17849 updated with directed reaction manually, without the help of equation mapping? |
Main improvements in this PR:
A total of 590 Human-GEM reactions were associated Rhea ID, and the corresponding master Rhea id. More information about the mapping process to Rhea can be found here
I hereby confirm that I have:
developas a target branch