Duplicate Beta-Oxidation Pathways for Linoleoyl- and gamma-Linolenoyl-CoA

[Devlin-Moyer]

Current behavior:

There are two separate pathways for the β-oxidation of linoleoyl and γ-linolenoyl-CoA to 4-decenoyl-CoA:

Overall Reaction	“Low-Resolution” Version	“High-Resolution” Version
γ-linolenoyl-CoA -> 4,7,10-hexadecatrienoyl-CoA	MAR05358	MAR03452 + MAR03453 + MAR03454 + MAR03455
4,7,10-hexadecatrienoyl-CoA -> 2,4,7,10-hexadecatetraenoyl-CoA	MAR05195	MAR03456
2,4,7,10-hexadecatetraenoyl-CoA -> 3,7,10-hexadecatrienoyl-CoA	MAR05302	MAR03457
3,7,10-hexadecatrienoyl-CoA -> 2,7,10-hexadecatrienoyl-CoA	MAR05229	MAR03458
2,7,10-hexadecatrienoyl-CoA -> 5,8-tetradecadienoyl-CoA	MAR05208	MAR03281 + MAR03282 + MAR03283
linoleoyl-CoA -> 7,10-hexadecadienoyl-CoA	MAR05356	MAR03275 + MAR03277 + MAR03278 + MAR03279
7,10-hexadecadienoyl-CoA -> 5,8-tetradecadienoyl-CoA	MAR05191	MAR03280 + MAR03281 + MAR03282 + MAR03283
5,8-tetradecadienoyl-CoA -> 3,6-dodecadienoyl-CoA	MAR05123	MAR03284 + MAR03285 + MAR03286 + MAR03287
3,6-dodecadienoyl-CoA -> 2,6-dodecadienoyl-CoA	MAR05075	MAR03288
2,6-dodecadienoyl-CoA -> 4-decenoyl-CoA	MAR05072	MAR03290 + MAR03292 + MAR03293

Metabolite	ID in “Low-Resolution” Pathway	ID in “High-Resolution” Pathway
4,7,10-hexadecatrienoyl-CoA	MAM03274m	MAM00091m
2,4,7,10-hexadecatetraenoyl-CoA	MAM03653m	MAM00072m
3,7,10-hexadecatrienoyl-CoA	MAM03221m	MAM00089m
2,7,10-hexadecatrienoyl-CoA	MAM03182m	MAM03009m
5,8-tetradecadienoyl-CoA	MAM03978m	MAM01576m
3,6-dodecadienoyl-CoA	MAM03203m	MAM01575m
2,6-dodecadienoyl-CoA	MAM02698m	MAM03021m
7,10-hexadecadienoyl-CoA	MAM03650m	MAM01577m

Expected behavior:

Both of these pathways should not exist; I propose removing the “lower-resolution” one and keeping the “higher-resolution” one. The pathway for β-oxidation of (4Z,7Z,10Z,13Z,16Z)-docosapentaenoyl-CoA (mentioned in #746) joins with the “lower-resolution” pathway at 4,7,10-hexadecatrienoyl-CoA; replacing MAM03274m with MAM00091m in MAR05360 will be sufficient to keep that pathway intact after removing the “lower-resolution” pathway.

Proposed changes:

• Remove MAR05358 for being a duplicate of MAR03452 + MAR03453 + MAR03454 + MAR03455
• Remove MAR05195 for being a duplicate of MAR03456
• Remove MAR05302 for being a duplicate of MAR03457
• Remove MAR05229 for being a duplicate of MAR03458
• Remove MAR05208 for being a duplicate of MAR03281 + MAR03282 + MAR03283
• Remove MAR05356 for being a duplicate of MAR03275 + MAR03277 + MAR03278 + MAR03279
• Remove MAR05191 for being a duplicate of MAR03280 + MAR03281 + MAR03282 + MAR03283
• Remove MAR05123 for being a duplicate of MAR03284 + MAR03285 + MAR03286 + MAR03287
• Remove MAR05075 for being a duplicate of MAR03288
• Remove MAR05072 for being a duplicate of MAR03290 + MAR03292 + MAR03293
• Replace MAM03274m with MAM00091m in MAR05360
• Remove MAM03274m for being a duplicate of MAM00091m
• Remove MAM03653m for being a duplicate of MAM00072m
• Remove MAM03221m for being a duplicate of MAM00089m
• Remove MAM03182m for being a duplicate of MAM03009m
• Remove MAM03978m for being a duplicate of MAM01576m
• Remove MAM03203m for being a duplicate of MAM01575m
• Remove MAM02698m for being a duplicate of MAM03021m
• Remove MAM03650m for being a duplicate of MAM01577m

[haowang-bioinfo]

this issue presents the duplicate pathways of beta-oxidation for linoleoyl- and gamma-Linolenoyl-CoA, which were from the rough merging from two sources, one from HMR2 while another from Recon3D.

the proposed removal of "low-resolution” version appears to be a reasonable solution, here are some ideas for implementation:

• better ignore potential inaccuracy of GPRs, and focus on reaction manipulation
• the subsystem name of kept reactions probably should be renamed
• the H2O2[m] in MAR03456 may be replaced with FADH2[m] from MAR05195

[Devlin-Moyer]

Re: changing the subsystems of the reactions not being removed: it looks like MAR03452, MAR03453, MAR03454, MAR03455, MAR03456, MAR03457, and MAR03458 (the reactions between gamma-linolenoyl-CoA and the point where its beta-oxidation pathway merges with linoleoyl-CoA's beta-oxidation pathway) are all in the Omega-6 fatty acid metabolism subsystem, while the rest are in the Beta oxidation of di-unsaturated fatty acids (n-6) (mitochondrial) subsystem. The reactions being removed are all in the Fatty acid oxidation subsystem.

There are only 19 reactions in the Beta oxidation of di-unsaturated fatty acids (n-6) (mitochondrial) subsystem, and nearly all of them are mentioned above. It seems redundant to have separate subsystems for Beta oxidation of di-unsaturated fatty acids (n-6) (mitochondrial) and Omega-6 fatty acid metabolism, since the former is a subset of the latter, and they're both pretty specific. I feel like it'd make sense to either

1. Reassign all reactions currently in the Beta oxidation of di-unsaturated fatty acids (n-6) (mitochondrial) subsystem to the Omega-6 fatty acid metabolism subsystem, or
2. Move MAR03452, MAR03453, MAR03454, MAR03455, MAR03456, MAR03457, and MAR03458 from the Omega-6 fatty acid metabolism subsystem to the Beta oxidation of di-unsaturated fatty acids (n-6) (mitochondrial) subsystem, and look through the other reactions in the Omega-6 fatty acid metabolism subsystem to see if they can also be reassigned to more specific subsystems, or
3. Add all reactions currently in the Beta oxidation of di-unsaturated fatty acids (n-6) (mitochondrial) subsystem to the Omega-6 fatty acid metabolism and Fatty acid oxidation subsystems, and add all reactions currently in the Omega-6 fatty acid metabolism subsystem to the Fatty acid oxidation subsystem without removing them from their current subsystems (i.e. have each reaction be in multiple subsystems simultaneously)

Depending on what the philosophy behind assigning reactions to subsystems is.

[feiranl]

During check, I also identified the duplicate mets in Peroxisome compartment, should we also fix this?

Metabolite	ID in “Low-Resolution” Pathway	ID in “High-Resolution” Pathway
4,7,10-hexadecatrienoyl-CoA	MAM03274x	MAM00091x
2,4,7,10-hexadecatetraenoyl-CoA	MAM03653x	MAM00072x
3,7,10-hexadecatrienoyl-CoA	MAM03221x	MAM00089x
2,7,10-hexadecatrienoyl-CoA	MAM03182x	MAM03009x
5,8-tetradecadienoyl-CoA	MAM03978x	MAM01576x
3,6-dodecadienoyl-CoA	MAM03203x	MAM01575x
2,6-dodecadienoyl-CoA	MAM02698x	MAM03021x
7,10-hexadecadienoyl-CoA	MAM03650x	MAM01577x

[Devlin-Moyer]

yea I also noticed that when putting together #793 and wasn't sure if it would be better to address that here or to do separate issues for duplicate peroxisomal pathways. I'd sorta tentatively decided to hold off on bringing up peroxisomal equivalents to the various mitochondrial duplicate pathways/reactions I recently identified until more of this current batch of issues had been resolved.

[feiranl]

Ok, I would say let's make it a separate issue later