I am new to this field and I have a conceptual question related to the ReactionDecoder tool.
I am running the tool in the Annotation mode. In the resulting XML file, I get multiple reaction centres in the form of substructures. From the EC-BLAST paper, I realised they are arranged in a fingerprinting pattern. I am testing this tool because I am interested in capturing the reaction centre of a particular reaction. What I am not sure is out of these multiple reaction centres which one should I choose, the 1st one (because the first fingerprint is closest to the reaction centre) or the longest one (because it is more explanatory) or should I consider all the substructures?
The reason I am interested in reaction centres is because I have a set of reactions where all the atoms of involved molecules are annotated with likelihood scores. A higher likelihood score for an atom means a possible site of metabolism (SOM). The likelihood scores are calculated using the recently published GNN-SOM approach. By capturing the reaction centre I also want to ensure the atoms within the reaction centre have higher likelihood scores for being a SOM compared to other atoms outside SOM.
Also, is there a document where I can get more information about the XML file tags? Because I see the resulting tags are dynamic depending on the type of reaction.
I am new to this field and I have a conceptual question related to the ReactionDecoder tool.
I am running the tool in the Annotation mode. In the resulting XML file, I get multiple reaction centres in the form of substructures. From the EC-BLAST paper, I realised they are arranged in a fingerprinting pattern. I am testing this tool because I am interested in capturing the reaction centre of a particular reaction. What I am not sure is out of these multiple reaction centres which one should I choose, the 1st one (because the first fingerprint is closest to the reaction centre) or the longest one (because it is more explanatory) or should I consider all the substructures?
The reason I am interested in reaction centres is because I have a set of reactions where all the atoms of involved molecules are annotated with likelihood scores. A higher likelihood score for an atom means a possible site of metabolism (SOM). The likelihood scores are calculated using the recently published GNN-SOM approach. By capturing the reaction centre I also want to ensure the atoms within the reaction centre have higher likelihood scores for being a SOM compared to other atoms outside SOM.
Also, is there a document where I can get more information about the XML file tags? Because I see the resulting tags are dynamic depending on the type of reaction.
Many thanks for your time.