fix: revise rxns and GPRs associated with DECR1 and DECR2

[haowang-bioinfo]

Main improvements in this PR:

This PR revise reactions and GPRs associated with DECR1 and DECR2

• remove MAR03322 and MAR03296 due to compartment cofactor mismatch, as reported in DECR1 and DECR2 use NADPH, not NADH #736
  • remove MAR04005, and associated MAM00678c and MAM03035c, due to compartment mismatch
• sort out GPRs according to compartment localication, as porposed in Localization of DECR1 and DECR2 #755
  • set the GPRs of MAR01176, MAR01181, MAR01218, MAR01222, MAR03446, and MAR03457 to ENSG00000104325
  • set the GPRs of MAR01297, MAR03461, and MAR03471 to ENSG00000242612

I hereby confirm that I have:

• Tested my code on my own computer for running the model
• Selected develop as a target branch
• Any removed reactions and metabolites have been moved to the corresponding deprecated identifier lists

[Devlin-Moyer]

All the changes look good to me; the only note I have is that in the PR message you say MAR03322 and MAR03296 are removed because of a "compartment mismatch", but the problem is that they used the wrong cofactor (NAD instead of NADP), not that they happened in the wrong compartment. Anyway I assume that was a typo, and since you link directly to the relevant issue where the cofactor situation is explained, it's not much of a problem.

[haowang-bioinfo]

All the changes look good to me; the only note I have is that in the PR message you say MAR03322 and MAR03296 are removed because of a "compartment mismatch", but the problem is that they used the wrong cofactor (NAD instead of NADP), not that they happened in the wrong compartment. Anyway I assume that was a typo, and since you link directly to the relevant issue where the cofactor situation is explained, it's not much of a problem.

sorry, that's my mistake and now corrected